( Reaffirmed 2006 ) Indian StMard APPLICATION GUIDE FOR INSULATION CO-ORDINATION ( First Revision) High Voltage Techniques Sectional Committee, Chairman San: V. R. NARASIMHAN Repwlding . ETDC 19 Central Electricity Authority, New Delhi to Members DEPUTY Dramxoa San: Shri V . R . Narasimhan ( Alkrnak ) Punjab State Electricity Board, Patiala K. S. Bix*anwAJ Sum A. K. CwoPaA ( Alkrnak ) Siemens India Ltd, Bombay Da S. C. BAATSA Da D. P. SAHGAL ( Alkrnak ) The Calcutta Electric Supply Corporation Da K . DAS GUPTA Calcutta Smar A. K. SERI v. BARMAN ( Ltd, Alkmak ( ) B. DESAI Jyoti Ltd, Vadodara Alkmatc ) Da P. SATYANARAYANA Smr V. S. MANX SKI D. V. Nnrure .%a1 B. N. Geosa Hindustan Brown Boveri Ltd, Bombay ) &RI K. S. MADAAVAN ( Alkmak .Smu S. K. MtxHEaJEE National Test House? Calcutta bharat Heavy Electncala Ltd, New Delhi ( Alkmak ) !&ax P. N. SRalvAsrAvA ( Alkmak ) The Bombay Electric Supply C Transport UnderSRRI V. H. NAVKAL taking, Bombay' Saat M. L. DoNoRE ( Alkmak ) Indian Institute of Science, Bangalore Pam G. R. GOWNDA RAJU DR B. I. GURURAJ ( Altemak ) Kamataka Electricity Board, Bangalore .%iRlK. S. SWAPRAKASAM Swru H. M. S. LXNQAIII ( Afkrnafe ) W. S. Insulators of India Ltd, Madras haI V. SamvAsAN Smu K. THIRUvENKADATXAN ( Alkmak ) SHar S. KR~PANIDHX( %WZAMANIAK Sxax T. V. Tamil Nadu Electricity Board, Madras Alknuk ) @ Cbpyrfght ,I983 fbis of 19S7 ) Rnd mprodu~ti~tt in whole or in part by my m~aarexcept with written pamissiot~ of tba pub&her nbaIl be deemed to be an infriorwot of copyripbt uodar the rid Act. BUREAU OF INDIAN STANDARDS publication is protected under the fn4m Copyrf& Acr (XIV fSr3716-1978 ( c5nriRwd~om pugc1 ) Members SHnI SURENDXASINQH Sam KOMAL SINGH ( Alfemak ) @menring U.P. Government Khurja Pottery Development Centre, Sanx c. R. Crompton Greava Ltd, Bombay VARIER DR G. PARTHASARATIIY ( Altwnafc ) The Tata Hydra-Electric Pow.er SUPPlY Co Ltd, SHRI P. J.* WADIA Bombay DR R. RANJAN ( Alkmah ) Director General, IS1 ( &~fi& Mnnbn) Srmx S. P. SACHDEV, Director ( Elec tech ) s#cY&qY Sxnr M. N. Amiint MURTHY Director ( Elec tesb ) 2 CONTENTS PAGE 0. FOREWORD I. SCOPB 2. TERMINOLOGY . . . . . . . . . . . . ..I . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . .. . 6 7 8 8 17 25 3. VOLTAGE STRESSESINSERVICE 4. INSULATIONWITHSTAND 5. PROTECTIVE DEVICES . . . .., 6. CO-ORDINATION BETWEENSTRESSES AND WITHSTANDCONSIDERA31 ... TIONS COMMON TO RANGESA, B AND C ... ... 7. CO-ORDINATION BETWEEN STRESSES AND WITHSTANDVOLTAGE ... ... . .. FORRANGE A .. . ... 8. CO-ORDINATION BETWEENSTRESSES AND WITHSTAND VOLTAGEFOR ... ... . .. . .. .. . RANGBB 9. ~~;~TION BETWEEN STRESSES AND WITHSTAND VOLTAGE ... ... ... ... FOR 1.. 33 43 43 APPBNDICES APPENDIX A APPENDIXB SURGETRANSFERENCE THROUGH TRANSFORMERS . . . 61 VALIDITY OF SWITCHINGAND LIGHTNING IMPULSE WITHSTAND TESTS, 50 PERCENT DISRUPTIVE DISCHARGE .. . TEST AND 15 IMPULSES WITHSTAND TEST . . . 72 STATISTICAL EVALUATION OFTHE PROTECI.IVE EFFECT ... OF SPARK GAPS . . . ... .. E_SAMPLES OF APPLICATION . . . ... ... 80 83 APPENDIXC APPENDIX D APPENDIXE CLEARANCES IN AIR BETWEEN LIVE CONDUCTIVE PARTS AND EARTHED STRUCTURESTO SECURE A SPECIFIEDIMPULSEWITHSTANDVOLTAGE FOR DRY ... .. . .. CONDITIONS .. . 87 TABLES TABLE 1 TABLE 2 PROVISIONAL SCALE OF NATURAL POLLUTION LSVELS PROVISIONAL .. . 32 33 RELATION BETWEEN POLLUTXONLEVPIS, . . . . . . TEST LBVBLSAND CREEPAOB DISTANCE 3 PAUR TABLU 3 CNOIOB OF THE INsuLAm LBVEL 01 CA1 I BQuxPHuaT (FOR &LE Tkmmmeua hBY fik?RoB . . . . . . a.. . . . I)IVSRTRRS) (I'CBRfiSAMPLRU- 84 OFTHEI~mum~NLava~ TABLB 4 CNOICB TABLE 5 op Chs~ IIl!&ume~~ OTRDDXSOONNROTINOS~)... Al A . . . hNCTIOX8 VOLTAOES OF TW OF THS 85 Rrsr:OFFAIL~RR OF hsu~&moN RATED %VITCHl NO hPUISE WlTHSTAND EQUIPWMT OF TABLR 4 . . . . . . 86 -_ TABUZ 6A ~RRfCLA~IONS BMWEXN TADLE 6B CoRRmmoss Manruy hWLATION ~~R?~R~~&A$E-T~-EARTHAIR~~ BETWMN PIUSE-TO-EARTB L.lWEts AND . . . AND 88 I~svnlron LEVELS Am CL.EARAN~S . .. 88 FIGURES Fxo. 1 FIG. 2 m. 3 O~RVOLT~~ES ... ... ... ... 18 22 42 46 49 PROBABIUTY OF DISRUFFIVE VOLTAGE UNDRRI~ hhXIMUM ARMSTBR htHIMBI.JZ OF LIN+ABIJE OF TXE R~R . .. CABXS op INSULATION ... hNC3TH ... WTPH &JR= . .. Jtnvorro~ ONLY . . . OF . . . PURER OF ... Fxca. 4 Fso. 5 EVALUATION k3ULATION !SIXPL~~~ FAILURE OF A ... ... STATISTICAL Msrmto~ . . . .. . FIG. 6 RxmorF~itune (8) AND STAGmm~~~~aBansoar TE62IOALm h&TC3X (y) Ff3RVARIOUS~W~~ SURGE DrmuewHoNs ... Rna OF ... ..* CORRRLA~RS I~TwRRN 51 Fxo. 7 FAUWUZ TISTIOAL~A~~YFA~~R soaot Drmmmmxom (y) FOR .. . V~IO~~SWMUUNG .. . . .. (R) AWD STA- 52 Fro.- 8 OFFAMJRE (R) AND STII CORRBLA~ONS BETWBEN RISK 'ITIRCALSAPE~Y FACTOR (y) FOR VAWWS Swrmmn, ... ... . .. SUR~B DmRIBuTIoNs 4 53 ISr3716.1976 PAQE FIG. 9 CORRELATIONS BETWEEN RISK OF FAILURE (R) AND STATISTICALSAFETY FACTOR (y) FOR VARIOUS LIOHTNING . . . . . . *.. Suxmu D~~~IBUTI~NS CORRELATIONSBEIWEEN RISKOF FAILURE (R) AND STAT~SI~CALSAFETY FACTOR (y) FOR VARIOUS LIGHTNING . .. . .. ... SURGE DISTRIBUTIONS CORRELATIOIUS BETWEXN RISKOF FAILURE (R) AND STAT~~CAL SAFXTY FACXOR(~) mxt VARIOUS LIGHTNING .. . . . . . . . SuRoBIhTR1zwTxoNs BLOCK-DIAORANOF INSULATION Co-ORDINATIONAND . .. ... I . .. DESIGN . .. INITIALCAPACITIVE VOLTAGL SPIKB VALUESOFFACWR ` I' .. . . . . . .. . . . ... 55 FIG. 10 56 FIG. 11 57 59 62 65 FIG. 12 FIG. 13 FIG. 14 FIG. 15 DEFINITIONOFTHE INSULATION STRENGTH OF A PIE~R AT TBB TIME t AS A Fmmm OF THE OF&WMXXUT . . . . . . . . . PARAMETER K . . . FREQUENCY WITHSTAND 73 FIG. 16 FIG. 17 DENST- OF THE MEAWRED 90 PERCENT w OF A POPWATI~N OFAPP;ARATW~... 75 PROBABILITY QBAN l!&IPMBNT To PASS TEIE DIFFEBEHT TYPE OF Tzsr AS A FUWXION OF ITS Iracwma; cnAltAmwrncs FIG. 18 FIO. 19 FIG. 20 thmomds hx . .. & M w A ... A . .. ..a .U ..* . . . . .. ... 76 77 78 79 PROBABiLXl'T~~OFT~~AlLURE P~TOFAXLTHSTEST thTOMER'8 ==Rm FUNCTION cm Ik-rI'Y ... I.. .*I ... hSK & FtnmrxoaOF bthNUFAm=R'S 3 lS:3716-1978 Indian Standard APPLICATION GUIDE FOR INSULATION CO-ORDINATION ( First Revision) 0. FOREWORD 0.1 This Indian Standard ( First Revision ) was adopted by the Indian Standards Institution on 19 April 1978, after the draft finalized by the High Voltage Techniques Sectional Committee had been approved by the Electrotechnical Division Council. 0.2 This standard was first issued in 1966 and covered recommended practices for the co-ordination of the insulation of electrical equipment located in electrically exposed situations. The insulation levels recommended in the 1966 edition were on the basis of IS: 2165-1962;. This revision has been undertaken with a view to bring this guide in line with the latest technological developments taking place in the field of insulation co-ordination like the switching overvoltages. This revision is aligned with the revision of IS : 2165 which has been issued in 1977 wherein proper emphasis on the switching overvoltages has been covered. 0.3 Probabilistic concepts and probabilistic language have also been introduced in this revision for the procedure of insulation co-ordlnatiori. This revision acknowledges that the engineers, particularly those who. work with extra high voltage field equipment, with the help of powerful computers, are now in a position to make use of such concepts, which afford a better knowledge of system and equipment behaviour and should contribute to a more economical design. 0.3.1 The traditional approach to insulation co-ordination was and still is, to evaluate the highest overvoltage to which an equipment may be submitted at a certain location on a system, and select from a table of standardized values the withstand voltage presenting a suitable safety margin. Both overvoltage evaluation and safety margin selection are largely empirical and, in many cases, the choice of the insulation level is still more readily based on the previous experience in the system or other similar systems. *Guide for insulation coordination. 6 IS : 3716- 1978 The object of this standard is not to give strict rules for insulation co-ordination and design, but to provide a guidance toward rational and economic solutions. Therefore, it is intended to consider only a few basic cases, it being evident that stations constituting exceptions to normal design, or included within systems having exceptional characteristics, will require special study by experienced engineers. 0.3.2 In a more elaborate process, it is recognized that over-voltages are random phenomena and that it is uneconomical to design plants with such a high degree of safety that they can sustain the most infrequent ones. It is also acknowledged that tests do not ascertain a withstand level with a 100 percent degree of confidence. In consequence, it is realized that insulation failures can occur occasionally in well-designed plant, and that the problem is to limit their frequency of occurrence to. the most economical value, taking into account equipment cost and service continuity. Insulation co-ordination should be more properly based upon an evaluation and limitation of the risk of failure than on the u priori choice of a safety margin. 0.4 In accordance with the latest decision taken at the international level, it has been decided to use the term `surge arrester' in place of `lightning arrester'. 0.5 In the preparation of this standard assistance has been derivedfrom IEC Pub 71-2 ( 1976 ) `Insulation co-ordination - Part 2 : Application guide' issued by the International Electrotechnical Commission. 0.6 A typical example for 220 kV transformer has been included in Appendix A to make the guide more useful since 220 kV system is more in vogue in our country. 0.7 For the purpose of deciding whether a particular requirement'of this standard is complied ivith, the final value, observed or calculated, expressing the result of a test, shall be rounded off in accordance with IS: 2-1960*. The number of significant places retained in the rounded off value should be the same as that of the specified value in this standard. 1. SCOPE 1.1 This standard provides guidance on the selection of the electric strength of equipment, of surge arresters or protective gaps, and of the most suitable degree of switching overvoltage dorrtrol. The rated withstand voltages of the equipment are covered in IS : 2165-1977t which forms a necessary adjunct to this standard. Non - This guide is based on apparatus types and ratings in use at present. As new equipment and equipment characteristics are developed and proved, this guide should not be interpreted as a limit to their adoption. *Rules for rounding off numerical values (mid). thdatiou co-ordination ( sucendretin ). 7 IS : 3716- 1978 \ 1.2 This'guide covers_only phase-to-earth insulation and deals separately with the three following ranges of the highest voltage for equipment: a) Range A : above 1 and less than 52 kV. b) Range B : from 52 to less than 300 kV. c) Range C : 300 kV and above. 1.3 It covers installations of all kinds and in all situations involving voltages higher than 1 kV, whether they are exposed to lightning or not, with the exception of overhead lines. However, the test procedures apply to the latter also. 2. TERMINOLOGY 2.1-For the purpose of this standard the definitions given in IS : 2165-1977* shall apply. 3. VOLTAGE STRESSES IN SERVICE 3.1 General - Dielectric stresses on insulation may be classified as follows: a) Power-frequency voltages under normal operating conditions, b) Temporary overvoltages, c) Switching overvoltages, and d) Lightning overvoltages. In IS : 2165-1977*, the overvoltages ar,n classified with reference to the shape of the voltage wave which determines their effect on insulation and on protective devices, without reference to the cause of the overvoltages. 3.1.1 The term `temporary overvoltages' refers to sustained overvoltages, or to overvoltages having several successive peaks, with a decrement of the amplitude such as to be comparable with a sustained voltage at power frequency or at harmonic frequency. 3.1.2 The term `lightning or switching overvoltages' refers to overvoltages for which only the highest peak value has to be considered and which can be represented, with regard to their effects on insulation and protective devices, by the steep front standard lightning impulse or the slow front standard switching impulse used for test purpose. The foregoing names have been chosen because such overvoltages often, but not always, originate from lightning discharges or switching operations. 3.1.3 For example, the energization of a transformer-terminated line gives rise to an overvoltage that may be regarded as a switching or temporary overvoltage depending on the decrement of the successive peaks ' &ulation co-ordination ( second m&ion ). 8 (that is, depending on the circuit constants). As another example, a lineto-earth fault, although actually a switching operation (the same phenomenon would arise if a phase conductor was connected to earth by a circuit-breaker operation) may give rise to steep-front overvoltages, similar to those due to lightning; on.the other hand a lightning surge transferred through a transformer by inductive coupling between windings may produce on the secondary side of the transt'oettd waves, similar to those due to switching operations. 3.2 Power Frequency Voltagea - In insulation co-ordination rocesSes, since overvoltages and impulse voltages are defined in terms oft Reir peak vah.uzs to earth, it is also convenient to make use ofthe phas&to-earth peak value of the system voltage, which is &/ q/3 I a816 times the usual rms phase-to-phase voltage. Under operating conditions, power-frequency voltagecan be expected to vary somewhat in magnitude and may be described by means of a probability distribution about the average operating value. This distribution will differ from one point of the system to another. For purpose of insulation design and co-ordination, it should, however, be considered as constant and equal to the highest voltage for eqmpment, which in Range C does not materially differ from the highest system voltage, with a peak phase-to-earth value of Urn d/2 1 d/3. In Range A and in Range B up to 72.5 kV the highest voltage for equipment may be substantially higher than the highest system voltage, as given in 2.3 of IS : 2165-1977*. For the sake of standardization it is, however, assumed that equipment insulation will always be .able to operate satisfactorily at the -highat voltage for equipment immediately above, if not equal to, the highest system voltage. 3.3 Temporary Overvoltages - The severity of temporary overvoltages is mainly characterized both by their ampIitude and duration. The importance of temporary overvoliages in insulation co-ordination is twofold: a) on the one hand the characteristics of temporary overvoltages at the surge arrester location are of great importance in surge arrester selection; and b) 011 the other hand, the successive repetition of overvoltage peaks of opposite polarity, even if of lower amplitude than soxix other overvoltages may determine the design of both the internal insulation of equipment as well as the external insulation ( surfaces exposed to contamination ) . *Insulationco-ordination ( smndratsbn). 9 I6 I 3716- 1970 Temporary overvoltages generally arise from: and a) earth faults, b) load rejection, c) resonance and ferro-resonance. 3.3.1 Earth Faults - The overvoltage at power frequency on the sound phases when another phase is accidentally earthed depends, at a given point of the system, on the treatment of the system neutral with respect to earth, as characterized by its earth fault factor at that point. 3.3.1.1 In the evaluation of the earth remarks should be considered : fault factors the following In general, in order to evaluate this factor at a given location, it is assumed for simpli@ty that the fault is located at the point for which the factor is desired; but, in some special cases, it may be desirable to investigate the effect of other locations on the highest value of the voltage to earth. In principle, there are as many particular values of the earth fault factor at a given location as different possible configurations of the system. The factor which characterizes the location is the highest of the values that correspond to the different system configurations which may occur in practice. The system configurations which have to be considered are those which exist during a fault; thus one should take into consideration those changes in the system which may be produced by the fault itself, for example, on account of the operation of circuitbreakers. For many systems, it will be sufficient to consider only one value of the earth fault factor which covers all the locations on the system. Attention is drawn to the fact that the highest voltage at system frequency which may appear on a sound phase during a particular earth fault does not depend only on the value of the earth fault factor but also on the value of the phase-to-phase voltage at the time of the fault. This phase-to-phase voltage will generally be taken at the highest system voltage, as given in 2.2 of IS : 21651977*; but, in some cases, in order to predict the operation it is of protective devices and specify their characteristics, necessary to take into account the increased value of the phase-to-phase voltage that may appear at the selected location under the abnormal conditions not covered by this definition. l I~ulationio-ordination ( secondrevision). 10 33.1.2 Within Range A and in some cases within range B many systems or installations are operated with their neutral earthed through a high impedance; an arc-suppression coil or with their neutral isolated. For the purpose of insulation co-ordination particular attention shall therefore, be paid in these cases to the earth fault factor. 3.3.13 Independently of the earth fault factor particularly high overvoltages may arise in Range A and Range B systems in the case of: a) earth faults in a system the neutral of which is earthed through an arc-suppression coil when the circuit is under-compensated. b) arcing earths in a system, the neutral of which is isolated and in many cases in a system the neutral of which is earthed through an arc-suppression coil. 3.3.2 Sudden Changes in Load - In the usual conditions of operation the phase-to-phase voltage does not exceed the highest voltage'of the system ( see 2.2 of IS : 2165-1977*); but higher values may temporarily be reached in the case of sudden disconnection of large active and reactive loads; they depend on the system layout after disconnection and on the characteristics of the sources (short-circuit power at the station, speed and voltage regulation of the generators, etc). This voltage rise may be specially important in case of load rejection at the remote end of a long line ( Ferranti effect ). It affects mainly the apparatus at the station connected on the line side of the circuit-breaker. the point of view of overvoltages, a distinction should be made between NOTE -From various types of system layouts. Extreme cases, namely, those with relatively ahort lines and high value of the short-circuit power at the terminal stations; and those with long lines and reduced value of the short-circuit power at the generating site may be considered. With the latter layouts, as are usual in an extra-high voltage system in its initial stage, much higher overvoltages at system frequency may result when a large Ioad is suddenly disconnected. Due to the characteristics of the systems, overvoltages of this kind are more severe in voltage Range C than in voltage Range B, overvoltages of this kind, in voltage Range A, may occur in generator transformer-circuits. 33.3 Resonanceand Ferro- Resonance -Temporary overvoltages due to these effects may generally arise when circuits with large capacitive elements ( lines, cables, series compensated lines ) and inductive elements ( transformers, shunt reactors) having non-linear magnetizing characteristics, arc energized. These situations are generally found for systems in Ranges C and B in the following cases: a ) A lightly loaded line, fed or teminated by a transformer may show, for example, harmonic oscillations and pronounced overvoltages if the natural frequency of the linear part of the system corresponds to one of the harmonics of the transformer magnetizing current. l &mllationco-ordination( MCOnd &tioll ). 11 b ) Subharmonic odlations and ovcrvoltagcs may occur in series compensated systems terminated by lightly loaded power transformers or shunt reactors if the impressed voltage, the effective circuit resistance, which is strongly influenced by synchronous machines,. and the `kircuit capaciknce fall between certain Ii&its. c ) If harmonic filters are connected to a system containing saturable elements, oscillations due to resonances between these elements and the filter capaciton can develop. These f-o-iesonance efTects following energization processes are either stationary or last several cycles of the power frequency being related to the time constant of transformer in-rush currents. 3.3.3.1 BTt: Causes of resonance and ferro-resonance in voltage Range A 4 resonance between inductive and capacitive components, for example, ivhen capacitors for power-factor correction are used. which may occur when a tran&rmer whose secondary is loaded by a small capacitance only, is switch& in or out, with an appreciable time between operatSons on each pk ferro-resonance which may occur where there. is an iron-cored inductive load such as a voltage transfokner ahzrdy connected or being switched in. md Ughtniag Ovrrvokagea W ferro-resonance 4 3.4 Switching 3.4.1 For the purpose of this guide switching overvoltages are, as stated above, of a m-which may be simulated by a standard switching impuk, ihat is, an aperiodic wave with a front duration of the order of humkda-qf microseconds and a tail the order of thousands .of microseconds. They stress the various parts of an insulation in about the same proportions as power-frequency voltages but arc not repetitive and only one peak of either polarity normally significant. of is 3.43 Lightning overvoltages are thosq which may be simulated by a standard lightning impulse, that is, an.aperiodic wave with a front duration the order of one microsecond and a tail durition of the order \of several Due to the front steepness they strew more than the tens df microseconds. former the longitudinal insulation inductive w, and because of their shorter duration, generally a somewhat higher stress may bi withstood by a given insulation. of of 3.4.3 These overvoltagea generally arise from: ia) line energization and raenergization, b) faults and fault clearing, 12 I3:3716-1978 4 swim 4 4 ofcapscitive currents and of small or moderate inductive current% loadrejtcti~and li&ing strokes (first or subsequent components of a lightning . 3.4.3.1 Linemrgidon and rs-snnrgiration oven&ages - Overvoltages due to closing, and to single and threophase reclosing are of great importance in the selection of system insulation in Range C. In this voltage ranp;mstrike-fke breakers are in general use. Oven&ages due to line energization and re-energization particular importance in the other ranges of voltages. are not of 3.4.3.2 Szvi&ag oruvoitag~ dut to faults and fault &wing - In Range A and, in relatively kw cases, in Range B under the conditions listed in items in 3.3.l.3 high switching overvoltages can arise at the &tLZr (o?! fault . In all the voltage ranges, high overvoltages may arise due to taults to earth in composite circuits (overhead lines and cables as transformer terminated f&n). At the highest voltages of Range C a high degree of control of overv&ages caused by Ike energizatiou and re-cnergieation is normally attempted. Forthis mason switching ovcrvoitagesdue to faults and firult &wing ( single and daubie liitiearth faults and their clearance) need cvehrl co&&ration. 3.4.3.3 &#&@ps dyr b sunkhing of hdwtivr ad cqgdeitioc currents In Range A the switching tif inductive or capacitive currents may give rise tie M w&h may raqt+e attei+n, both in high voltage d&t+-. b&r& *~*$&$a ;rp$ ia i* m$&&&o ns and power stationa. % the casic 4 L e timer, high overvoltagea may ark if the c&u&breaker deiakea so rapidly ti to f&ma &a current prematu$y t.9 26x0, so-called current+pping. a) Iritaruption of thr Wig currenzs of motor& b) kee _diaductk cummt, for example, when .interrupting ZW@Et=lngcwFeatOfatr~~Or~Or; cr Sting and operation of arc funuces and their tranakmers vlhich may lead to current chopping; 13 J8 : 3716 - 1978 d) Switching of unloaded cables and capacitor banks; and e) Interruption of currents by high-voltage fuses. In Range B overvoltages due to the interruption of capacitive currents (switching off unloaded lines, cables or capacitor banks ) may be particularly dangerous since the use of restrike-free breakers may not alway be assumed. 3.4.3.4 Load rejection - Overvoltages due to load rejection may start with a high switching surge followed by a temporary overvoltage. Overvoltages of this kind are particularly important in Rang C at the highest voltages where a high degree of control of reclosing surges is attempted. 3.4.3.5 Lightning overvoltages - Lightning overvoltages are caused either by direct strokes to the phase conductors, back-flashovers, or as a result of earth flashes in the proximity of the line which produce indirect lightning surges. The overvoltages by which subs:ation insulation is stressed are a The confifunction of the line construction and the system configuration. guration of the station itself has a great influence if the travelling time of surges within the station is not negligible in relation to the front time of the surge. Depending on the system configuration, overvoltages with time parameters in the range of switching surges may also arise as a result of lightning strokes. Lightning discharges which produce significant overvoltages, in Range B and C, are confined to direct strokes to phase conductors or strokes to towers or earth wires with subsequent back-flashovers. In Range A indirect lightning surges shall also be considered. Furthermore in this range surges transferred through transformers from a higher voltage system need careful consideration. 3.3 Determination of the Expected Overvohage Level 3.5.1 Range A -For voltages of less than 52 kV switching overvoltages constitute generally no serious problem for overhead supply systems and insulation co-ordination is based on lightning overvoltages. Switching overvoltages transferred from an overhead line into a plant through transformers or lengths of cable-may, in general, be ignored for connected to the same reason. An exception is the case of an insdation the lower-voltage side of a high-voltage transformer feeder, particularly if resonance occurs between the two systems during ane or two-phase energizing. IS t 3716 - 1978 In industrial plants and power stations, the amplitudes and wavcshapes of switching overvoltages generated within the installation vary over a very wide range. In the great majority of cases they are innocuous; in some, serious overvoltage magnitudes and rates of change may occur. Thus sudden voltage swings may be caused when a switching device re-strikes; the resulting rate-of-change of voltage may equal that caused by a severe close lightning strike. A very large amount of practical operational experience is available from Werent industrial plants and power stations and the most severe overvoltages or voltage swings may usually be avoided by eliminating resonance and by the correct choice of switching device to be employed. Detailed representation of the system under consideration on the digital computer or on the TNA may not be economically justified at this voltage level since accurate representation is needed to obtain accurate results and a complex plant frequently consists of many pieces ofequipment and electrical connections. Furthermore the operation of some types of switching device and arcing earths arc difficult to simulate with a sufficient degree of accuracy. Experience is often the -best guide and in exceptional cases, deliberate switching tests with simultaneous recording ( both highspeed and low-speed ) will produce the most valuable information so that remedialmeasures may be taken as the result of subsequent calculations and confirmatory tests. The amplitudes, waveshapcs and frequency of occurrence of lightning overvoltages on systems in Range A may be estimated with a reasonable degree of accuracy. As the impulse flashover voltage of insulators used on overhead lines in this rauge is quite low as compared with the potential impressed on such a line by a direct lightning stroke, the stresses to which substation quipment is liable to be subjected are primarily determined Thus .careful protection of substation by the type of line construction. equipment is required if this is connected to a wood- ole line with unearthed crossarms. Reduced protection is needed where Jl e lines are erected on steel masts, reinforced concrete poles or where metal crossarms are otherwise earthed. Apart from this important difference, the amplitude and waveshapes are affected by the following factors which characterize the constitution of the system and arrangement of the station. 35.1.1 Surge im_b&nce of those lines or ccblts which are conneclrd to thr - For example, when only one line is connected to a terminal transiz the surge is reflected at the termination and is doubled in voltage amplitude; when n lines of the same surge impedance are connected to the busbars of ziitation and if no lightning stroke to the line occurs near the station, the voltage at the busbars becomes 2 u/n, where u is the amplitude of the surge voltage transmitted along the line on which the lightning surge originated. ' 15 I8 i 3716 - 1978 35.12? Cables with an earthed metallic sheath in serus with the line or connected between the station bwbars and apparatus to be protected - A short cable mainly reduces the steepness of most of the waves entering the station; a cable of In one or a few kilometrcs length may also reduce the surge amplitude. the GIX of a direct lightning stroke to the last span in front of a station, a cable section between' overhead line and station affords practically no relief to the station equipment. 35.13 Protective earth wires on the overhead lines extending up to a fm k&n&es ahead of the station - These are effective against close lightning This presupposes that strikes to the line which are the most dangerous. shielding by the earth wires is sufficiently well designed to prevent direct strokes reaching the phase conductors, and that the earthing resistance the tower is sufliciently low to reduce the risk of back flashover, for example, 10 ohms for a 36 kV line. of 35.1 .d Protective spark gaps or protective earth wires extending over one OT two front of the sta1ion - These may materially reduce the amplitudes of incoming surges on lines with high insulation to earth, for example, /on fully insulated wood-pole lines. spans in 3.5.1.5 Earthing resistances and i&lances of the down leads of towaz, particulars ciose to the station - In the cases of high values of earthing resistance or the inductance of the down lead of the tower or pole, a lightning strike to such a tower or pole or to an earth wire may cause high overvoltages on the phase conductors by back flashover across the line insulators to one or more phase conductors. 3.5.1.6 In Range A lightning surges transferred through transformer. and are also important. Analytical expressions for the electrostatic electromagnetic terms of the transferred voltage are derived in Appendix A. In this range of voltage4 also the insulation levels are 3.52 Range Bgenerally such that switching overvoltages are seldom a major problem and that insulation co-ordination is still mainly based upon lightning overvoltages in overhead line systems. Furthermore, also in this voltage range, there is usually no decisive economic incentive towards a detailed study of overvoltage stresses. Thus the considerations in 3.5.1 also apply to Range B. 3.5.3 ZZmrge C- Although a switching impulse test has been substituted in this range, as more realistic, for the traditional one minute power frequency test, it is only for the highest values of the range that switching overvoltaga become the predomiit factor in insulation co-ordination. The high cost of equipment then compels consideration of less liberal designs of insulation co-ordination, while, in turn, the serious comequenca of a failure necessitate a more precise estimation of the ovcrvoltageo `16 IS : 3716- 1978 These have to be evaluated for each type of significant to be expected. overvoltage in the particular system considered. Because of the extensive computational requirements, virtually all practical overvoltage predictions must be made using transient network analyzers or digital computers. Experience with studies of a wide variety of Fystcms has shown that development of generalized formulae for expected overvoltages is difficult because of the large number of parameters affecting the overvoltage value. Both analogue and digital techniques of transient solution require a reasonably high'level of skill in problem-solving. These skills are principally useful in the selection of significant cases ( it being impractical to study all possibilities ), in the reduction of the system to a reasonable number of busbars and lines ( it is not practical to represent the entire system on either TNA or digital solutions ) and in the description of system constants and apparatus characteristics. Whenever possible, field tests to check the validity of the parameters used are recommended. In the sophisticated approaches to insulation coordination now which are increasingly used for the highest values of voltage, the amplitudes of the overvoltages to be expected at a given location due to a given type of event, may not be defined by a single value ( see Fig. 1 ). It is only possible to state what is the probability f, ( U) dU that an overvoltage value comprised between U and U + dU may occur, f. ( U ) being the overvoltage probability density. The probability Fo ( U' ) that the value U' may be exceeded is then given by: m Fo(u> = 4. INSULATION 4.1 General U' s ~(U)dU'jir`F,(U')= -Jo 5 U' ( U ) dU. . ..(l) WITHSTAND 4.1.1 Self-R&wing and .Non-Self- Restoring Insulation - Clauses 2.8 and 2.9 of IS : 2\6,65- 1977' subdivide insulation into self-restoring and non-selfrestorin8insulation according to its behaviour in case of the occurrence of a disruptive discharge during a dielectric test. On the former kind of insulation it is possible to carry dut tests under conditions that imply an appreciable risk of such discharges, for example, by applying a large number of impulses at the rated impulse withstand voltage, or even in conditions with deliberately applied discharges as in a 50-percent disruptive dipcharge test carried out at voltages above the rated impulse withstand level. l xnmll8tion co-oKlh8tion ( umd rmisior, ). I7 lS : 3716 - 1978 IA Line Encrgization Overvoltages IEi overvoltages Across an Insulator String Due to Lightning Strokes to the Tower f. ( IJ 2 = overvoltage probability density me ( u ) = overvoltage 1;:~. 1 & ( cumulative ) probability ERVOLTAGXS 18 IS : 3716 - l!B% On non-self-restoring insulation a disruptive discharge destroys the insulating property ,of the insulation and a large number of impulses at rated withstand voltage may result in a gradual deterioration of the insulation. Non-self-restoring insulation is for these reasons tested by application of a limited number of impulses at rated withstand voltage. The degree of information on the dielectric strength of the equipment directly obtainable may thus be much higher for self-restoring insulation. However, in the case of non-self-restoring insulation, the economic importance for the manufacturer of the risk of having the equipment rejected tends to oblige him to design the equipment for a very low probability of failure under test. Taking these two factors together, no difference is made in IS: 2165 - 1977* between impulse withstand levels, in relation to the kind of insulation or the nature of the test. While self-restoring insulation does not lose or modify its insulating ability following a disruptive discharge in a dielectric test, it should not be inferred that damage may not occur in service if the disruptive discharge is followed by an intense power arc. Furthermore, possible damage to equipment is not the only consideration to be introduced in the selection of an acceptable risk of discharge in service, as the effect on continuity of supply also has to be considered. For example, a much lower probability of insulation failure is required in the case of bus-bars than on Ldividual lines. It shall be emphasized that the insulating structures of a piece of equipment are always made up of self-restoring and non-self-restoring parts. Generally it may not, therefore, be stated that the insulation of an apparatus is self-restoring or non-self-restoring. But the probability that discharges may occur across or through non-self-restoring parts irk the presence of selfrestoring rarts may, for different types of equipment, be negligible or not. Due to the different voltage-time discharge characteristics of solid and air insulations, this probability tends tc increase with increasing impulse voltage amplitudes: thus it may be negligible at the rated withstand voltage but may become appreciable around the 50 percent disruptive discharge voltage. 4.1.2 SeLxtion cf the Type of Test- For some types of apparatus, within the range of overvoltages that tests have to simulate, the probability that a In this cask discharge occurs across a non-self-restoring part is negligible. the discharge probability coincides with that of the self-restoring parts of the apparatus and its insulation may be called essentially self-restoring; or, for the sake of simplicity, self-restoring. Disconnecting switches may be considered an example of this type; in fact even when applying impulses well above the 50 percent discharge voltage during a 50 percent discharge test, sparkover takes place usually in air without any puncture of the porce*Insulation co-ordination( scrond revision ). 19 IS : 3716 - 1978 lain. For this type of equipment is possible and recommended. the test given in 7.3 of IS : 21651977* Insulation of other pieces of equipment, for instance bushings, behaves like self-restoring insulation up to the rated impulsS withstand voltage level, but this implies an overinsulation of the non-self-restoring parts as compared with the self-restoring ones, and this overinsulation, although economically permissible, hasto be limited to an acceptable extent. Thus, above this value a non-negligible probability exists that discharge takes place on non-self-restoring parts. Insulation of this type of equipment, that could be called combined, should be tested according to 7.4 of IS : 21651977*. Finally, the cost of the non-self-restoring parts of some equipment, for instance power transformers, may be so high as to make overinsulation unacceptable. Insulation of this kind of equipment is called essentially non-self-restoring or, for the sake of simplicity, non-self-restoring. This kind of insulation is verified by means of the test described in 7.5 of IS : 2165-1977*. Considerations Appendix B. on the validity of the above tests are given in 4.2 Insulation Behaviour at Power-Frequency Voltage and Temporary Overvoltages-In general, discharge under power-frequency voltage in normal operating conditions and under temporary overvoltages will be caused by progressive deterioration of the insr:lating properties of the equipment or by exceptional reductions in insulation withstand due to severe ambient conditions. In the latter case, the concept of ccntamination ( see 3.2 ). of probability is applicable to the degree Because of the difficulties involved, no use of statistical concepts will be made in these specifications in respect of insulation behaviour at power-frequency voltage and temporary overvoltages ( see also 6.1 and 6.2 ). of Discharge of Insulation Under Impulse VoltagesThe ability a given insulation to withstand the dielectric stresses caused by the application of an impulse of given waveshape and peak value U is, in most cases, a random phenomenon, even if w-e consider a time interval so small ( such as that needed to carry out a dielectric test on equipment ) that the ambient and insulation conditions may be considered constant, at least in respect of quantities such as pressure, temperature, humidity, etc, which mav be measured and which are used to define the ambient and insulation'conditions during tests. -i The discharge probability of an insulation to an impulse of given waveshape and polarity, and to aspeak value U in a short time interval as 4.3 Probability of l Emulation co-ordination ( secot.~! mision ). 20 IS:3316-1978 defined above ( for example, in a dielectric test ) may be determined, if the insulation ia self-restoring, by applying the impulse U `.iV' times within this time interval, and counting the number `n' of discharges. From the value $ a numerical value may be obtained for this probability which will be-the more accurate the greater the value is of `N'. Determination of the discharge probability of a given piece of nonself-restoring insulation caused by the application of a very small number of voltage impulses of given waveshape and peak value U, which shall be withstood without any failure is, obviously not possible; but this should not prevent us from regarding it as existing in theory and studies are continuing on the probability aspect of non-self-restoring insulation ,failure. If we consider either switching or lightning impulses of different peak values U, we shall be able to associate with every possible value of U a discharge probability Pt, thus establishing a relationship P, ( U) for a given insulation in a short time interval at or for the sake of simplicity at a time t ( see Fig. 2A ) . The values of P, ( U) increase from near 0 to near 100 percent probability in a more or less narrow band of voltage values. In general the resulting curve may be defined by a biparametric law, one parameter being associated with the position of the band and giving an indication of the withstand level, and the other associated with the bandwidth and giving an indication of the scattering of the voltage values which give appreciable proportions of both discharges and non-discharges. Generally in a laboratory the parameter that defines the position of the probability curve is taken as the voltage Ut,, which corrcspJnds to the 50 percent discharge (withstand ) probability. The standard deviation of the distribution (ct ), which corresponds to half the difference between the voltages that give discharge probabilities of 16 percent and 84 percent, is usually taken as the parameter which expresses the scattering. In the service the ambient and insulation conditions do not remain constant. Therefore, the discharge probability curve of insulation, as defind above for the time t, is bound to change from one moment to another (P,, , P t"... . ..) ( see Fig. 2A ). The variations are determined, as regards external insulation, mainly by atmospheric conditions. Taking the ambient and insulation conditions as random, it will be necessary to consider for each insulation, in addition to the discharge robability of probability P, ( U), as defined above, also a discharge insulation PT ( U) to overvoltages of amplitude U liable to !? occur at any For the purpose of insulainstant of a long time interval T of operation. tion design, it is this second distribution which is of interest to the engineer ( see Fig. 2C >. 21 , r IS:3716 -1978 (O/o) 100 04 16 0 ?.A utso "i J< u u; 2B (%I 50 u 100 84 50 16 0 UT50 2c U FIQ. 2 PROBABILITY OF DISRUPTIVE DISCHARGE ( FOR GAUSSIAN DISTRIBUTION) OF INSULATION UNDER IMPULSE VOLTAGE 22 rs 13716- 1978 Similarly to P, ( U),. Pr ( U ) nmy be. d&necI by the voltage Urss which corr onds to the 50,percent discharge ( withstand ) probability and by the stan ?z rd deviation er of the distribution. The variations of Pt( U) within the time interval AT may be conveniently defined by the probability density Pn which is a function of VW taken as a random variable ( scc Fig. 2B ). This latter function in turn may be characterized by the 50 percent discharge voltage uTjo and by its standard deviation an. On the simplifying assumption that the standard deviation ut of Pt ( u) is constant within the time interval Ar, the following relation hoti aT= V a$+ an' -(2) In clause 2.22 of IS: 21651977* the parameter that defines the p&ion of the probability curves P (v) is taken as the voltage which corresponds to a withstand probability of 90 percent although the 50 percent discharge voltage which was referrcd to above is a convenient measure for pieces'of insulation that may be submitted to a 50 percent disruptive discharge tests. The reason for this choice is that the 50 percent disruptive dischagc test may not generally be applied to all kinds ofinsulation. Thus, in order to have the same value of the rated impulse withstand voltages for all types of equipment, whatever its insulation, and to use these values directly in the definitions of the statistical distributions, it has been deemed appropriate to refix to a higher value ( 90 percent) of the withstand probability, the rated impulse withstand voltage URw being identical with the lowest permissible value of the statistical impulse withstand voltage under specified test conditions ( Ut,)). For risk-of-failure evaluations it is, however, convenient to express the probability curves of insulation discharge in terms of their 50 percent discharge voltages and standard deviations. Assuming for Pt (U) a Gaussian distribution with standard deviation et, the difference between the 59 percenh e&barge ( withstand ) vohgc and the statistical or 90 percent wi&stand w is given by: I+=1 &e l-3 ar . . . (3) at is taken as PO3 or Q-66 dependin & the type of impulse, lightning or switching, unlessanother value has %een speci&d for the relevant equipment (see 7.3 of ISt2165-1977*). The probability of discharge Pt ( U) of a piece of equipment which during the test is confbrrning to what is specified in IS: 2165-1977. may l IesuWon ~Miation ( rrcond rmirion )- 23 PIt3716-1376 then be defined in terms of its 50 percent standard deviation, as follows: discharge URW voltage and its Go - 1.3 ot - 1.3 ut I.. (4) The Pt ( U), defined by the above parameters, refers to the moat severe test conditions for the equipment, since URw is the rated switching or lightning impulse wrthstand voltage. Therefore, if impulse tests have to be made with the equipment both dry and wet, Pt ( U) refers generally to the wet conditions. The probability of discharge in service PT ( U) of a given piece of equipment may be deduced from field tests only, depending on the site of installation. Iiowever, as a broad indication the PT ( U) of a piece of equipment conforming to what is specified in IS : 2165-1977* may bedeflned, recalling equation 2, in terms of its 50 percent discharge voltage and its standard deviation, as follows: where k is the ratio between the 50 percent discharge voltage of a given equipment in service during a time interval A r and the 50 percent discharge voltage under the most severe impulse test for the equipment ( wet or dry, positive or negative polarity ). For switching impulses U of positive polarity, the values ofk and an relevant to time intervals A I of fine dry weather or various bad weather conditions do not show appreciable differences. The same may be said concerning the degree of ambient pollutron, at least in the range from clean conditions to lightly polluted conditions. For switching impulses U of negative polarity, the values of k and on are highly dependent on the type of weather within the time interGal Ar under consideration. Concomitance of bad weather (rain, snow, fog, mist etc) and not negligible pollution leads to a low value ofk; bad weather also increases the value of on. Values of k - 1 and un = 5 percent are suggested in these recommendations for normal conditions and a time interval Ar equal to the seasonal cycle to cover the worst polarity impulse, This value of aa results in a value of or a little lower than 8 percent. *Jnrulation co-ordination (second n&ion j. 24 I!%:371611376 The same values of A and an are also suggested for lightning This gives a value of eT equal to 6 percent approximately. impulses. The information given above is to be considered merely as broadly indicative and it is recommended that use shall be made of more detailed data derived from field tests if available. equipment with high-voltage 4.4 Equipment with Windings -An windings, such as a transformer or reactor designed to withstand only fullwave tests, is vulnerable to a certain extent to a surge of high amplitude chopped in its vicinity, because higher internal stresses than under full-wave All flashovers conditions may be developed across adjacent turns and coils. to earth in a substation result in chopped waves of various degree ofampliIf because of the use of protective spark gaps these are tude and steepness. liable to occur frequently in service the strength of the windings against The surges shall be determined by testing with a suitable chopped wave. provision for such a test is left to the relevant equipment specification. Where non-linear resistor-type surge arresters are protection of transformers, chopped waves are less likely chopped wave tests are usually not required. used for the to arise and For all types of apparatus having windings, such as rotating machines, transformers and reactors, rapidly changing voltages due to the restriking of switching devices may also produce non-linear voltage distributions similar to those caused by lightning overvoltages. For this reason, it is recommended that such equipment, irrespective of whether or not it is to be used in installations subjected to lightning overvoltages, should be tested with a lightning-impulse voltage to check the winding insulation for voltage withstand across turns and coils. 5. PROTECTIVE 5.1 General DEVICES devices fall under three classes: surge arresters; ( Range A only) ; and surge arresters These a) non-linear c) spark gaps. resistor-type b) expulsion-type The choice between these three devices, which do not provide the same degree of protection, depends on various factors, for example, the importance of the equipment to be protected, the consequences of an interruption of service, etc. In the following clauses, the point of view of insulation their characteristics co-ordination. will be considered from 5.2 Non-linear Resistor-Type Surge Arresters - These protective devices should be designed and installed to limit the magnitudes of over25 voltages against which they have to protect equipment so that the total surge-arrester voltage during operation does not exceed an acceptable value. They are defined and their characteristics are given in IS: 3070 ( Part I )-1974*. Their rating is defined as the designated maximum permissible rms value of power-frequency voltage between their terminals at which they are designed to operate corre~fly; this voltage may be applied to surge arresters continuously without changing their operating characteristics. In addition to this defined capability, some types of surge arresterst may successfullywithstand either (a) higher than rated voltage for a specified short duration or (b) a specified number of successive discharges. In either case a controlli factor in the selection of the surgearrester s ollow current at either the rated voltage or is its ability to interrupt power at the higher temporary overvoltages. A primary poitit is that the total voltage produced across the terminals of the arrester at any moment during operation shall be considered in the determination of the switching impulse protective level and the lightning impulse protective level. 5.2.1 Lightning Impulse Protective Lwel - The lightning impulse protective level of a surge arrester is characterized by the following voltages: a) The sparkover voltage for a standard full lightning impulse wave [ seeTable 3 of IS : 3070 ( Part I)-1974; 1; b) The residual ( discharge ) voltage at the selected standard nominal current [ seeTable 4 of IS : 3070 ( Part I )-1974* J; and c) `i'he front-of-wave sparkover voltage (Part I )-1974; 1. [ seeTable 3 of IS : 3070 Nary.-- The tables mentioned here give for each surge arrester voltage rating the upper limit for each of the above voltages. If better characteristics than those zqcc&d in IS : SO70 ( Part I )-1974* are available, the actual voltages for the specific surge arrester will be obtainable from the mannfWurcr. `IX?. it is recommended that the actual voltages for the surge arrester protective characteMtia be used for co-ordination studic3. The protective level under lightning impulses is taken for insulation co-or&nation purposes as the highest value among either (a), (b) or (c) divided by l-15 ( see 2.29 of IS: 2165-1977x j. This evaluation of-the protective level gives a conventional value representing a generally acceptable approximation. For a better definition of wave-front protection by a surge arrester, reference should be made to IS : 3070 ( Part I )-1974*. *Specification for lightning arresters for alternating current systems: Part I Non-linear tTbcu qxcial types of surge arresters arc at present applicable to Range C only. $Insnlation co-ordination ( secuad r&a ). resistor type lightning arresters (first n&.&a ). 26 I6 t 3716 - 1!978 5 9.2 Switching Impulse Prot&vc Level - The switching impulse protective level of a surge arrester is charauterized by the following voltages: _ a) The maximum sparkover voltage for the standard impulse shapes specified in 7.7.4 of IS : 3070 ( Part I )-1974*; and b) The total surge arrester voltage exhibited by the surge arrester when discharging switching surges. The protective level for switching impulses is the higher value of (a) or (b). Until astandard test for (b) is devised, reference should be made to the surge arrester manufacturer. 5.5 Espdsion-Type Surge Arresters - These protective devices operate to limit overvoltages and interrupt follow currents within their rating. They have low residual voltages. The characteristics of these devices are given in IS : 3070 ( Part II )-1966t. The impulse sparkover characteristics resemble those of protective spark gaps but are in general lower and flatter for the same sparkover distance. These arresters may not appreciably limit the amplitude of the follow current before interrupting it and may have current-interrupting ratings which must be compared with the prospective fault current and the prospective transient recovery voltage at the point of installation. 5.4 Spark Gaps - The spark gap is a surge protective device which consists of an open air gap between an energized electrode and an earth electrode. On supply systems operating at voltages up to 245 kV spark gaps have proved satisfactory in practice in some countries with moderate lightning The adjustment of the gap settmgs shall often be a compromise g"eFCei: perfect protection and service continuity but this difficulty may be largely overcome by the use of rapid automatic reclosing. wentidly The sparkover voltage and the time-to-sparkover of the gap depends on the distance between the electrodes; they are influenced by the shape of the electrodes and also by their disposition and distance r&the to the neighbouring live and earthed parts. In order to improve the operation of a spark gap under steep-fionted surges and to provide a flatter impulse sparkover-voltage time characteristic, the geometrical conf&uration of the simple rod to rod electrode arrangement may be mod&d, for instance by appropriate shaping of the electrodes and l Spccifieation for lightning arresters for alternating eurrcnt systems: Part I Non-linear m&or type lightning arredtcn (Jrd ~&&II ). $3p&kation for lightning arrcatcrs for alternating current ryrtcms : Part II Ex@aio~ type lightning arresters. 27 I8:3716-1978 by provision of a central auxiliary electrode. In Range A duplex-type gaps have also proved advantageous in regions where birds or small animals are troublesome. 5.4.1 Protective C%amctit&s of u @ark Gap -The protection obtained by means of spark gaps is less precise and the protective level may not be given as precisely as the protective level of non-linear surge arresters for the following ,reasons: a) The dispersion of the sparkover voltage of a gap; and b) The increase in the sparkover voltage with increasing amplitude of the applied wave when sparkover takes place on the front of the wave. The performance of a gap under impulse (switching or lightning ) is characterized by the 50 percent value and the standard deviation of its Since spark gaps discharge voltage under standard laboratory conditions. constitute typical self-restoring insulation, the contents of 4.3 apply to them as well. Furthermore, because of the reasons given in item (b) above, knowledge of the times-to-sparkover of the gap for values of the applied impulses w& aboyt the 50 percent sparkover value is often needed (see tive Gaps 5.4.2.1 W'hen the.gl$-operates on a voltage surge and a power-arc results, it frequently persists until disconnected by a fault protective device; this constitutes a short-circuit in the case of a system with directly earthed neutral, entails mechanical stresses on the various parts of the system equipment and may cause disturbances to consumers. The location of the gap should therefore, be considered in relation to its effect on the system protection and operation. 5.4.2.2 The gap is unacceptable from the point of view of service countinuity if its presence noticeably increases the number of circuit outages provided these flashovers are neither self-extinguishing nor interrupted by means of high-speed tripping followed by high-speed reclosing. 5.4.2.3 Spark gap operation causes chopping of the wave, thus increasing the probability of producing chopped waves close to the termiThis has to be taken into consideration for nals of protected apparatus. insulation of high voltage windings ( see 4.4 ). 5.4.2.4 Damage to the apparatus may be caused by the power arc across the gap if this is not installed in a suitable position. For instance if a spark gap is fitted to a bushing, for example, of a transformer or circuitbreaker, its distance from the bushing surface shall be sufficiently large to prevent a power arc being blown against the insulator, 28 Is t 3716- 1378 5.5 Application of DiGrent Protoctivc Devices 5.5.1 Protection with .Non-linear Resistor-Type Surge Arresters - In order to ensure that an apparatus is not subjected to a surge voltage exceeding that which appears across a surge arrester, it is a general rule t,> locate the arrester as close as possible to the apparatus. In particular, surge arresters should preferably be either installed on the transformer tank or its high-voltage and earth terminals should be connected to the transformer by the shortest possible connections. Similarly surge arrc&s should be fitted close to cable terminations, if they need protection, with the shortest possible connections between the terminals of the surge arrester and the phase conductor and the cable sheath respectively. NOTE - Jn the case of surge arresters close to the apparatur following conventional safety factors are recommended: to be protected the a) Range -4 -A safety factor of approximately 1.4 should be provided between the rated lightning impulse withstand level of the apparatus to be protected and the impuhe protective level of the surge arrester. b) Rongcz B and C- Conventional for lightning overvoltages. c) RcrngcC-Conventional switching ovcrvoltagcs. safety factors of I.2 - 1.4 are normally provided of 1.1 - 1.2 are normally provided for safety factors The account. considerations in IS : Qc)04-1978. shall also be taken into The installation of surge arresters close to the apparatus to be protected may be achieved more easily in Range A than in Ranges B and C. When the surge arrester is separated from the apparatus to be protected the apparatus is subjected to a surge voltage which exceeds the protective level of the arrester. The excess voltage is due firstly to the inductive voltage drop in the connecting leads of the arrester itself and those connecting it to the apparatus to be protected. Secondly, if the time of surge propagation between arrester and apparatus is not negligible compared with the front duration of the incoming surge, the effect is a short-time increase of the voltage at the terminals of the apparatus to bc protected over the protective level of the surge arrester. The increase due to both these factors depends on a number of conditions, namely, distance of the surge arrester and its location ahead of, or behind, the apparatus to be protected; characteristics of the !ine;.capacimnce of the apparatus to bs protected; arrangement of the station and This increase may be limited by all steepness of the incoming wave. arrangemeut.s which limit the steepness of the surge arrivmg at the station *Application guide for non-linear systems ( second rcGsion ) . resistor-type surge arresters for alternating current 29 ISr37l6-1978 [ extension of shielding wires, local&d capacitance, cable ( even short ) large number of connected lines]. The adoption of a reduced protective level is another help. 5.5.2 Protection with Expulsion-ljpe Surge Arresters - These amaters are sometimes used on high-voltage distribution circuits where shielding against lightning is not provided ( Range A ). The impulse sparkover voltage time characteristics of such an arrester is flatter than that of a rod gap of the same sparkover distance, but not quite as flat as that of soAM type of equipment, for example, a transformer winding or a cable. For this reason, an adequate margin of safety is required, not only for the lightning impulse sparkover voltage of the arrester and of the equipment to be protected but also for the corresponding front-of-wave sparkover voltage. These conditions are assisted by the usual practice of installing these types of diverter close to the equipment to be protected. For further specification concerning the applications of these devices, reference should be made to IS : 3070 ( Part II )-l!Wj*. 5.5.3 Protection with Spark GapsThe impulse sparkover voltage-time curve of a spark gap is usually much more curved than those of some of the types of apparatus to be protected, particularly those of transformers and cables. Due to the curved shape of the voltage-time characteristic of a spark gap, the distance over which protection is given for all surgesis very small, usually not more than a few metres. If a spark gap is applied for protection against surges of a limited front steepness ( considerably lower than the steepness of the standard lightning-impulse test voltage wave), a distance of several tens of metres between the gap and the object to be protected does not appreciably modify the conditions for the protection provided again& such surges. A spark gap is, therefore, liable to operate not infrequently when stressed by lightning surges, and occasionally when stressed by switching surges, the amplitudes of which are below the lightning-impulse withstand voltages of the apparatus to be protected. In a large number of cases the operation ofthe spark gap causes a circuit outage if the gap is on the supply side of the opening switch. If the supply may be restored quickly by high-speed automatic reclosing the setr;ng of the spark gap may be so adjusted m to provide an acceptable degree of protection to the apparatus without causing an excessive number of troublesome supply interruptions to consumers. Nom- Salty factors of the order of thoJt giVm for surge arrcdtcn secure ge~ally satisfactory protection provided the oeeurrencc of very steep fronted rurga is cx&dcd ( Appendix c 1. arruterS for alternating current rystems *Specification for lightning :ype lightning am&em : Part II Expulsion 1 30 IS : 3716 - 1978 CtxumDfNAmN CONSIDERATIONS BammEzu COMMON sTB%ssES AND WITHSTAND TO RANGkS A, B anil C 6.1 hsah&a Design to Power lkqmmcy Upewing Voltage and Tempmaq Overvekages-Pa of Palm amd AgeingIn 3.4 and 35 of IS : 2165-1977* it has been specified that the relevant equipment specScation shall prescribe the long duration power-frequency tests intended to demonstrate the behaviour of equipment with respect to internal insulation ageing or to external pollution. Only general guidance is given to the equipment committees; it is indicated that as regards the voltage under normal operating conditions, the insulation shall withstand permanent operation at the highest voltage for equipment. 6.2 PoUutiam 6.2.1 For insulation susceptible to contamination, the problem ofspecif$ng a suitable test method and pollution severity levels is at present under consideration for various relevant equipment specifications. When contamination tests are established, it is anticipated that the system engineer will specify a degree of pollution severity level in relation' to the pollution of the ambient in which the equipment is installed. 6.2% Table 1 gives a provisional basis to the system engineer for establishing a qualitative degree of pollution severity. 6.2.3 A scale defined in quantitative terms with reference to a test method should be associated with each of the qualitative levels ofnatural pollution severity for various type of insulators. 6.2.4 Besides being reproducible, a test method should, as far as possible, satisfy the requirement of validity, that is, d a satisfactory simulation of the natural conditions in which the equipment is to be in:talled. Therefore, the most satisfactory tests, among those presently adopted, may vary from case to case . 6.2.5 It shall be stressed that Table 1 does not cover some environment at situations such as desert areas, where long dry perio.ds are followed by condensation or light rain. 6.2.6 As an example, for line insulators, rather than a specific apparatus, Table 2 gives an indication of the possible requirements corresponding to the various types of tests. - 6.2.7 An indication of the required creepage distance is also given, although it is recognized that the performance of surface insulation is greatly affected by insulator shape. 6.2.8 The data in Table 2 are intended to cover the behaviour of equipment at the appropriate voltage, that is, either UrJJ/J3 or U, in case of a system which may operate with a phase earthed for long durations. lInsadatiorl co-ordination( second rmhion ). 31 II) I 3716 - 1978 TABLE I PROVISIONAL SCALE OF NATURAL POLLUTION LEVELS ( Clauses 6.2.2 and 6.2.5 ) POLLUTION LavaL ENVIRONMENT PERPORMANCE OF EXISTING LINEs (3) No faults are observed in high humidity conditions ( fog, mist, etc ) on 145 kV lines even when equipped with less than 9 to 10 insulators of the normal type*, nor on 245 kV lines even when equipped with less than 15 such insulators. (1) No significant pollution (2! Areas without industries and with low density of houses equipped with heating plants; areas with some density of industries or houses hut subjected to frequent winds and/ or rainfalls. Ail areas shall be situated far from the sea or at a high altitude, and shall in any case not be exposed to winds from the sea. Areas with industries not producing particularly polluting smokes and/or with average density of houses equipped with heating plants; areas with high density of houses and/or industries but subjected to frequent clean winds and/or rainfalls; areas exposed to winds from the sea but not too close to the coast ( at least about 1 km). Areas with high density of industries and suburbs of large cities with high density of heating plants producing pollution; areas close to the sea or in any relatively case exposed to strong winds from the sea. Light Faults occur in fog conditions on 143 kV lines with leas than 9 to 10 insulators of the normal type* and on 245 kV lines equipped with less than 15 such insulators. Heavy Faults occur in foe conditions. or when the wind blows from the sea, on HV lines equipped with normal-type insulators* ( unless the number of units per string is exceptionally large : more than 11 to 12 units on 145 kV lines and more than 18 units on 245 kV lines). Faults occur in fog conditions or during salt-storms on HV lines, even when quipped with antipollution-type insulatorst ( unless the number of units per string is exceptionally high; more than 1 l-12 antipollution units on 145 kV lines and more than 18 antipollution units on 245 kV lines ) . V-y Herrvyt Areas, generally of moderate extension, subjected to industrial smokes producing particularly thick conductive deposits; areas, generally of moderate extension, very close to the coast and exposed to very strong and polluting winds from the sea. the following characteristic spacing: *Reference is made to normal-type ins~tlators with -__ -146 mm: diameter : 255 mm, creepage dutance : XXI mm. tAreas of moderate extension very close to highways where a mixture of salt and bitumen may cause severe deposits on the insulators may be subjected to a high pollution level. $The reference to antipohutiofftype i.mulators is somewhat vague. due to the great variety ofantipolhrtion-type insulators whrch are praently in servrce on HV lines. 32 Is a3716 - 1978 TABLE 2 PROVISIONAL RELATION BETWE& POLLUTION TEST LEVELS AND CREEPAGE DISTANGE ( Clauses6.2.6 and 6.2.8 ) POLLUnON LEVEL IJZVRX.S, TYPE OF TEST r-`-__--Salt Fog Method Solid LayerMethod7 Withstand Salinity Layer Conductivity (2) kg/m8 No sign&ant pollution Light Heavy Very heavy (5 10-20 40-60 2 160 (3) (PI 5-10 10-20 20-40 >50 (4) (cmlkV_1 2.0-2.5 30-95 4.0-90 >6 and pin insulating For other types of insulator and pawcularly for very large~insulators in substations, the correlation with present test methods, between test I* and the creepage distance and sexvice experience is not yet sticient to give more defimte indication& variour pollutions leveb and do not necessarily agree with the creepage distances mved NOTEI - The value reportedin the table were cs@blished on the basis of normal cap NOTES - The creepage diitances given in the table are those recommended f& ihe from thccol'3 of Table 1 which refer to existing lines whose behaviour to power frerlucncy voltage may or may not be satisfactory. 65.9 If temporary overvoltages are frequent and severe, it may be necessary to take them into account in prescribing the pollution test. 6.2.10 In the case of stations with a high degree of pollution, when it may be impossible, or extremely expensive, to ask for the necessary performance of equipment under pollution coziitions, the alternatives of greasing OI washing the insulating surfaces.should be considered. For insulation susceptible to ageing, the problem 6.3 Ageiagspecifying suitable test methods are also at present under consideration. 7. CO-ORDINATION BETWEEN VOLTAGE FOR RANGE A STRESSES AND -TAND of . 7.1 Selection of the Rated Power FreqnesU?y WfthM8ad VO&agO Table 1 of IS : 2165-1977* indicates for each system voltage &, one value of rated Gower-frequency withstand voltage only. *Insulationco-ordination ( &ond fruition ). .,,33 l8 I 3716 - 1978 7.2 Selection of the Rated Lightning-Impulse Withstand VoltageTable 1 of IS : 2165-1977* leaves open the choice between two corresponding rated lightning-impulse withstand voltages, according to List 1 and List 2. Reduced rated lightning-impulse withstand voltages have been used with good results and for a wide range of equipment over long periods of time. Comprehensive tests have also been performed on different types of equipment for this voltage range to determine their impulse withstand voltagesi both for standard lightning impulses and representative switching impluses. It has been found, in particular, that the breakdown voltage of insulation under typical switching impulses is always higher than that of the peak of the power-frequency test voltage. This is one of the reasons why it was not found necessary to introduce a separate switching-impulse withstand in Range A. The choice between List I and List 2 is to be made in accordance with 4.2 of IS : 2165-19778 and the following considerations relevant to the equipment installations : 4 Equipment with no connection to an overhead line, bl Equipment connected to an overhead line through a transformer, and Equipment connected to an overhead line either directly or 4 through a cable. 73.1 Equipment with No Connection to an Overhead Line-A wide variety of installations is covered by this category, for example large underground cable networks in cities, many industrial installations, power Equipment in such positions is not subject stations and ship installations. to any external ( lightning ) overvoltages but may be subjected to switching overvoltages ( se6 3.4.3.3 ). In 4.2 of IS : 2165-1977*, the conditions are specified under item 1 in which equipment to List 1 may be used in such installations. In all other cases, equipment to List 2 should be used and, with few exceptions, no mrge protection is required. One such exception is an electric arcfurnace installation where high overvoltages are liable to develop due to current chopping by a circuit-breaker. Protection by special surge arresters may be required in such a case both between phases and between phases and earth. 7.2.2 F&ulpment Traosfmer Connected to an .Overhead Line Through a 7.2.2.1 Gene*1 conrlderatlona - Equipment connected to the lower-voltage side of a transformer the higher voltage, side of which is supplied from an overhead line is not directly subjected to lightning or *Insulation co-ordination ( sswnd revision ). 94 IS : 3716 - 1978 switching overvoltages originating on the overhead line. However, due to electrostatic and electromagnetic transference of such overvoltages from the higher-voltage winding to the lower-voltage winding of the transformer, such equipment may be subjected to overvoltages which, in certain circumstances, may exceed its breakdown voltage. Analytical expressions for the electrostatic and electromagnetic of the transferred voltage are given in Appendix A. terms For a given transformer the magnitudes and waveshapes of these transferred overvoltages are mainly dependent on the nature of the lowervoltage circuit and, for this rewon, it is convenient to consider the selection of the rated lightning-impulse withstand voltage of the equipment and its protection separately for the two basic categories of installation as follows: Category I : Equipment connected through transformers to highervoltage overhead lines, and incorporating connections of moderate length, say up to 100 m, between the lowervoltage side of the transformer and the equipment, such as the main supply switchgear of a cable distribution network or an industrial installation. C&gory 2 : Generator-transformer 7.2.2.2 Basic guidance installation. a) Category 1 equipment - Factors which tend to increase the magnitude of transferred overvoltages for such equipment are: a transformer having a high voltage ratio and high capacitance betwan windings; 2) a transformer disconnected from its load on the lower-voltage side; low-capacitance connections between a transformer and its 3) associated equipment; 4) a higher-voltage winding which is not earthed ( for example delta or unearthed star ), or having a star point which is earthed through a high reactance ( for example arc-suppression coil ) ; 1) 5) surges having steep wavefronts and surges having long durations; and 6) Switching surges due to energizing a transformer from a remote point on an overhead-line system ( that is energizing a transformer feeder ). Estimates of the_ magnitudes of transferred overvoltages may be made and methods of calculation, with examples, are described in Appendix A. 35 IS : 3716- 1978 Category 1 equipment may usually be protected by surge arresters, and where such protection, is provided as a normal practice it is not necessary to make these calculations. For other cases, basic guidance is given below on the nature of the tranrferred voltages, the general influence of circuit ~~nditi~n~ and the criteria which may be used' to determine whether precautions are necessary. pared with the voltage to earth due to a surge on one phase of an overhead line, t&e phase-twphase voltage may theoretically approach twice as much for the same 8urgc or three times as much for equal surges of opposite polarity on two &arcs. Consideration should be given to the probability of this occurring in service and hence the precautions which are necessary to PmteCt againstit. Nola2 -For resortant voltagea, a condition of resonance between two systems connected by a transfbrmer may cause abnormally large voltpga to It is recommended that an examjnaby transfrmd through the trasx+forma. ticm of the circuits for pol$Mt resonance should be made, and modifications shobld be made as necessary to avoid WSXIUIC~. NOTBI -For rurge voltages between phases, transferred aurge voltaga Corn-bchvecn phases are often higher than those between phase and earth. The application of short duration, or steeply rising, voltage surges to the higher-voltage side of the transformer, for example, a lightning stroke to the transmission line very close to the transformer may through capacitive coupling, give a short duration voltage `spike' on the lower-voltage side. This may exceed the impulse test voltages given inTable 1 of IS : 2165-1977*. On the other hand, the %%test possible front time, determinedby wave impedance of the' line and transformer input capacitance, is often so long that the capacitively transferred voltage may be ignored. For Category 1 equipment the most onerous condition rises when the load circuits are disconnected, that is, on the transformer side of the lower-voltage switcbgear, since with the load connected its capacitance is ufl;Lafy suflicient ts~ reduce the amplitude of the initial voltage `spike' to a safe value. If the capacitance of the cosmeetionsbetween the transformer and lower-voltage switchgear is not sufficient to reduce the amplitude of the initial voltage `spike', either additional capacitance may be connected between tbe transformer terminab and earth, or equipment in accordance with Table 1, List 2 of IS : 2165-1977* shall be wed. It may also be desirable to consider adding surge arresters. Attention is also drawn to the possibility of increasing inductively trPnsfared overvoltagcs by additional capacitance. This increase l IilWktioa co-o&nation ( SocQDd r& ). 36 IS : 3716- 1978 may be reduced by a series damping resistor of carefully adjusted resistance value. The application of a longer duration voltage surge to the higher-voltage side of the transformer, for example, lightning stroke to the transmission line some distance from the transformer or a switching surge, will, through inductive coupling, give a voltage surge on the lower-voltage side of the transformer having a longer duration and an amplitude which shall be compared with the peak of the power-frequency test voltage given in Table 1 of IS: 2165-1977*. Dangerously high over-voltages can be transferred to the lower-voltage side of a transformer through capacitive coupling from the higher-voltage winding when an earth fault exists on the higher-voltage system and when the neutral of the system is earthed through an arc-suppression coil or when it is isolated. W Recommendations on the need for surge Catego~ 2 r&mreat-protection of generator-transformer installations and on suitable types of protective equipment need to be based on consideration of overvoltages of atmospheric origin only since studies have not revealed any more onerous condition likely to arise from Pransfenmce of switching surges. Corresponding to the front of an incident lightning surge or to the collapse of voltage due to chop ping there may be a capacitively transferred voltage of short duration ( initial voltage `spike' ). This is independent of the longer-duration voltage which is usually transferred by the combined effect of inductive and capacitive couplings. The maximum amplitude of the initial voltage `spike' is highly dependent on details of the design of the installation. Where these are such as to assist capacitive transference there may be justification for making a low-voltage surge-injection test on the installation or on the generator-transformer connected to a circuit simulating the generator and its external connections. Factors which tend to increase the magnitude of transferred over-voltages for such equipment are : 1) high capacitance between the transformer windings, connections between transformer and 2) low-capacitance generator, 3) high voltage r&o of transformer, to a 4) 3 Iamzr-voltage transfbrmcr winding not cnw generator, and 5) surgia having steep wavefronts and surges having iong duration. l rmBul8tioll co-ordin8tion ( d ftiti ). 37 IS13716-1978 If there are indications that the amplitude of the initial voltage spike should be reduced, this may be done effectively by connecting capacitors from each phase to earth by means of lowinductance connections, preferably at the lower-voltage terminals of the transformer. Attention is drawn, however, to the possibility of increasing inductively-transferred overvoltages by additional capacitors. The longer-duration transference generally takes the form of a unidirectional voltage with superimposed oscillations having a frequency of several kilohertz and if reduction of this is necessary, consideration should be given to the addition of surge arresters. However, voltage division between the reactances of the generator transformer and the generator normally ensures that the amplitude of the longer-duration transference does not warrant the use of surge arresters. If the generator transformer can be energized from the high-voltage system when the generator is disconnected, this voltage division does not occur and consideration should be given to the higher amplitude of the longer-duration transference affecting that part of the lowerFor voltage circuit which remains connected to the transformer. large generator installations, surge arresters located on the generator are not considered as protecting the low voltage side of a connected transformer and calculations should be made. In so far as surge arresters can readily be applied to small installations there is no necessity in these cases for calculations of transference to be made. The effects of the application of longer duration voltage surges to the higher-voltage side of a transformer and their transference to the lower-voltage side when an earth fault exists on the higher-voltage system when the neutral of that system is earthed through an arc-suppression coil, or if it is isolated, are subject to the same considerations as described for Category 1 equipment. 7.2.2.3 Selection of insulation level - The choice of whether to use List 1 or 2 of Table 1 of IS : 2165-1977* and whether additional overvoltage protection is necessary should be based in the first place on service experience with similar installations. It may also be useful to make measurements on an existing similar installation, using a low-voltage impuheinjection method. For a large installation, and where the necessary data concerning the transformer and the protective equipment are available, it will be useful to calculate the overvoltages liable to be transferred and to compare the results with the appropriate withstand voltages of the equipment to be *Insulationco-ordination ( sued mui* ). 38 IS t 3716 - 1978 protected. This is normally advisable only for direct connections between generator and transformer and for low-voltage tertiary windings on large system transformers. If a circuit-breaker is installed between a generatortransformer and its associated generator, consideration should be given to the cases when the breaker is closed and when it is open, although a load is usually connected to the lower-voltage side of the transformer by which transferred voltages may be reduced even in the latter case. Several methods of calculation have been published and, on the whole, these seem to give similar results. Although no absolute accuracy may be claimed for any method of calculation, comparison between calculation and experimental results on a variety of installations has shown satisfactory agreement. It is therefore deemed appropriate to illustrate a method- of calculation by reference to two numerical examples, covering respectively Categories 1 and 2. Examples are given in A-2. 7.2.3 Equi/munt Connected Directly to an Overhead Line - Equipment installed in a substation connected directly to an overhead line is subject to direct or indirect lightning overvoltages. Such equipment should, as a general rule, comply with the rated lightning-impulse withstand voltages specified in List 2 (Table 1 ) of IS : 2165-1977*. All equipment and, in particular, transformers in such positions require protection by surge arresters or spark gaps. Having regard to the flat impulse-breakdown time characteristic of a transformer winding, transformers should preferably be protected by non-linear resistor-type surge arresters in areas of intense lightning activity. In areas of moderate lightning activity, expulsion-type surge arresters may be used. Where lightning activity is slight, protective spark gapiZave proved adequate, particularly where the transformer is connected to a line with earthed crossarms or where the transformer is designed to withstand steep-fronted chopped waves. The bushings of circuit-breakers, instrument transformers and substation insulators having curved impulse flashover voltage time characteristics, may be effectively protected by existing protective devices on the transformers. In areas of moderate or low lightning activity, equipment having rated lightning-impulse withstand voltages in accordance with List 1 of Table 1 of IS : 2165-1977' may be used but, in that case, careful attention has to be paid to adequate covervoltage protection. In systems the neutral of which is earthed through a low resistance, surge arresters or spark gaps may be used for this purpose. In systems the neutral of_which is earthed through an arc-suppression coil, adequate overvoltage protection shall be provided; if surge arresters are used, those which can withstand repeated operations during the persistence of arcing grounds are recommended. l IslsulUion co-ordination ( UC& wvi#i8a). 39 I8 t 3716 - 1978 In the absence of any overvoltage-protective device, lightning surges impressed on an overhead line are limited only by sparkovers on the line at the weakest points which the surges meet during their `propagation. If not correctly localized, such sparkoven may cause damage to equipment as a result of surge r&ections between the point of sparkover and a vulnerable apparatus, such as a transformer winding. In the case of a substation with a number of lines normally connected to the busbars, the surge voltage arising at the busbars is likely to be sticiently reduced (stt 3.5.1 ) not to overstress apparatus in the station. However, such a solution ( no overvoltage-protective device ) may be acceptable in practice on overhead supply systems in regions of very low lightning activity, at least if equipment to List 2 of Table 1 of IS : 21651977*, is used. 7.24 Equij?mtnf Conntcted to an Ovtrhtad Line Through a Cable - Insulation ' co-ordination in this case is not only concerned with the protection of the substation equipment but also with that of the cable. In this respect the two ends of the cable may be assumed to be subjected approximately to the same overvoltage amplitudes. When a lightning surge propagated along an overhead line impinges `on a cable, the latter acts, substantially, like __a capacir'ance. Thus the front steepness of the original surge is reduced as the surge voltage enters the cable. The amplitude of Us of the surge entering the cable is given by: where u, = amplitude of surge voltage on overhead line; = surge impedance of overhead line, in practice & = 400 Zl to 500 ohms; and ,& = surge impedance of cable; in practice & = 25 to 50 ohms but for some types of cable it may be as low as 5 ohms. This initial surge is reflected at the station end of the cable in accordance with the effective surge impedance at the station busbar. Subsequent refleetions along the cable continue to be governed bv the above equation with due regard to the fact that VI and & invakably refer to the wave which impinges on a point of reflection while U, and 2s refer to the reflected wave. *Insulation co-ordination ( smnd r&.kn ). 40 I6 I 3716 - 1978 Provided at least one further cable of a few hundred metres `length is permanently connected to the busbar, the surge voltage to which the cable and the station equipment aresubjected is notably lower than that on the line on which the surge originated and this reduction is all the greater, the lower the surge impedance of the cable. However, if the cable only supplies a terminal transformer the As a result of incident wave is doubled in amplitude at the transfbrmer. successive reflections at both cable terminations, this voltage build-up increases towards twice the amplitude until no further energy is supplied by the original surge. For a station to which at least two cables are connected, a decision on the adequacy of equipment to List 1 of Table 1 of IS : 2l65-1977* or the need for overvoltage protection may be determined from the above equation. However, in the case of a terminal station, the ultimate surge-voltage amplitudes developed at the cable terminations as a result of successive reflections are a function of the amplitude and duration of the original lightning-surge voltage on the line, the length of the cable, and if the stroke is fairly close to the cable, also the reflections from the point of strike. For lines with fully insulated crossarm s, the resulting voltage ampli- tudes are so high that, even using substation equipment and a cable with lightning-impulse withstand voltages to List 2 of Table 1 of IS : 2165-1977+, surge arresters shall be used at the line/cable junction. As an example, the maximum cable lengths are plotted in Fig. 3 for which the cable and the substation equipment may be protected by surge .arresters at the line/ cable junction only; the figure demonstrates the considerable benefit of-the The protection is fully effective use of a cable of low surge impedance. against direct and indirect lightning surges and surges due to back flashover provided these originate a few spans away from the line/cable If the cable length exceeds the values indicated in Fig. 3 terminations. additional surge arresters are required at the substation end of the cable. If surge arresters with lower sparkover voltages than specified in IS : 3070 ( Part I )-1974t are used the cable lengths indicated in Fig. 3 may be increased in proportion to the differences indicated by comparing say a 10.5 kV with a 12 kV surge arrester. For lines with earthed crossarms feeding a cable with terminal transformer the impulse flashover voltage to earth of the line insulation i; only slightly hi her than the corresponding value in List 2 of Table 1 of IS : 2165-1977 # . In such a case, surge arresters may be required at the linelcable junction and it may also be necessary to use these at the station end of the cable. *Insulation co-ordination ( s8cond rroirion ). *Lightningarreatcn for altcmatiug eumzntsystems: Part I Non-linearr&or lightningarresters. 41 types `! 5 - c8blc surge inlpcdaMxin ohmr L - cable length in metrea hmp: flmum~~ FIO. 3 ~&UUMUM PERMISSIBLE CABUS LENOTIS WITIX OF LINE/CABLE JUNC~ON ONLY The foregoing considerations away from the cable termination. strokes is generally not possible. apply to direct stroker a few spana Full protection against very close In areas of moderate or low lightning activity, protective spark gapa may be used in place of surge arresters. However, if the spark gaps at the line/cable junction are earthed through a low resistance ( the usual case ) and if the cable is terminated in a transformer, dangerous surge voltages can be developed across the transformer winding. The spark gaps at the line/cable junction should therefore be earthed through a resistance of 42 IS : 3716 - 1978 several tens of ohms, ideally equalling the surge impedance of the cable. Greatly improved protection may be achieved by installing additional spark gaps-across he line insulators on the first and second poles in front of the line/cable termination and, in this case, the earthing resistances of these additional spark gaps are immaterial. 8. CO-ORDINATION BETWEEN VOLTAGE FOR RANGE B STRESSES AND WITHSTAND Voltage and 8.1 Selection of the Rated Power Frequency Withstand the Rated Lightning Impulse Withstand Voltage 8.1.1 Many considerations concerning voltages in Rahge A still apply to Range B. However, the variety of equipment and situations is not so great as in Range A. 8.1.2 In Table 2 of IS : 2165-1977*, one value of rated lightning-impulse withstand voltage only is associated with each value of rated powerfrequency withstand voltage. Therefore, there will be a unique choice for the rated power-frequency withstand voltage and the rated lightning impulse withstand voltage. 8.1.3 With each value of U, are associated one to three values of rated power-frequency withstand voltage with a corresponding rated lightningimpulse withstand voltage. 8.1.4 The choice between 72.5 kV shall take account of: the possible alternatives for U, above a) the neutral earthing conditions; and b) the existence of protective devices, their characteristics and their distance from the equipment considered. 8.1.5 The conventional safety factors normally employed in the application of surge arresters in Range B are given in 5.5& 9. CO-ORDINATION BETWEEN VOLTAGE FOR RANGE C STRESSES AND WITHSTAND $I Insulation Design to Power-Frequency Voltage and Temporary `Overvoltages - For this range of voltage, power-frequency tests are to be specified in the relevant equipment specifications, in accordance with the considerations in 6, and taking intoaccount that the temporary phaseto-earth overvoltages will not usually exceed 1.5 p.u. ( per unit ) for one second on each occasion. IS: $716 - 1978 9.2 Insdation Dcsigp to Switchin and Li&tninS OvervoltagesIS: 2165-1977* proposes two mcth J s for coordination of insulation in respect of switching and lightning overvoltages, namely, a conventional and a statistical method. 9.2.1 Conurational Method - This is based upon the established concepts of maximum overvoltages stressing insulation and of minimum strength of the insulation (2.20 and 2.23 of IS : 2165-1977* ). The statement of the minimum strength and that of the maximum overvoltages are rather arbitrary since a rigorous rule may seldom be followed in the evaluation of the upper and lower limits of the insulation strength and overvoltage value, which are intrinsically random variables. Insulation is s'elected in such a way as to achieve a sufficient margin between the maximum overvoltage and the minimum strength. This margin is intended to cover the uncertainties of the designer in the evaluation of the maximum overvoltage and of the minimum strength, and no endeavour is made to assess quantitatively the risk that in&&ion may break down. The conventional safety factors normally employed in the application of surge arresters in Range C are to be found in 55.1. 9.2.2 Stutirticuf Method - This attempts to quantify the risk of fhilure for use as a safety index in insulation design. Rational insulation design of a transmission system should be based on the minimum of the installation cost plus the capitalized yearly operational cost and yearly cost of failure, the latter being calculated as the estimated cost of failure of insulation multiplied by the average expected number of insulation failures per year. In order to evaluate the average expected number of failures per year of a piece of insulation located at a given point of the system in consequence of overvoltages, all the events giving rise to overvoltages whieh Then tbr may affect insulation design should be taken into consideration. each type of event considered the frequency of occurrence during the year and a separate diitribution of the overvoltage amplitude would be required. It b evidqt that the amplituda of all the overvoltaga occurring in 8 NFayater# can not be combined in one distribution but that only overvoltagea identified by the same location and cause may bc considered as statistically hemogeneow. Actudy, since the ovcrvc+gc severity differa for waveshapes whichare compar@le impulseand with a switching impulse ( w Note under tively with a ligh tnmV 977. ) the kervoltage &:F&d 2.18 of IS: 2165amplitude could be eaid to he h0m-u only if i&Mied by the same location, cause and shape. However the overvoltrga due to the lsune es*= at a given location`have broadly a imik shape %l8ulation co-ordktion ( su#nd & ). Is ; 3716 - 1976 same cause and lotition may, for the rake of al4ttlcnfaclhouidalti6edbytlu If problems of rtandardization of the UVdPsimpkity,be~alhomogalann. mcnt of an attire network are to bc dult with, ~II extion of the concept ofa *luJmogencouSgcuqJof oxrvoltagunceda to lx considered. In tbii caac a vouP of ovcrvoltagcS may bc said to be haarogmcous if the 0~er~01tages occur in Similar locationa of the q&m due to the same arue For acam~le the rcclosiog ovcrvoltages on the bus-but ( sending-end ) of any cobjtation of tbc q~tem may be considered `a a bomogcneo~ gtoup ofo==ltag= However, a process of insulation design and coordination as outlined above involves too many difficulties. The statistical approach considered here is thvefbre restricted to checking that the risk of insulation failure due to any foreseeable type of event causing overvoltages in the system is within acceptable limits. These limits depend on the frequency of occurrence of the type of event and on the consequences of the failure of the piece of insulation under consideration. Fortunately, the types of event which are significant in insulation design are generally a few in number to allow an analytical approach. For instance, the insulation withstand to switching surges of many pieces of equipment. in a system is, in general, determined primarily by reclosing overvoltages. If the frequency distribution of overvoltages caused by a given type of event and the corresponding insulation strength are known, thi risk of failure ma-? be expressed numerically, as will be shown below. Let the withstand strength of a given piece of insulation within a given time interval Al be defined by the probability PT ( U) of discharge of the insulation when it is subjected to an overvoltage of value U ( sac Fig. 4 j. Furthermore let the distribution of the over-voltages stressing the same piece of insulation for the specific type of event considered be defined by rhe probability density fO( U). Then the probability that an ~~LQ!&+Q~ of value comprised between U' and U' + dU may occur is fo( C' ) d4f: %e'probab&tv density-of i&:urc of the insulation due to an overvoltage value U' `is, therefore, the product of the probabihty dens* that an overvoltage of value u may occur and the probability that the insulation may fail under an overvoltage of value U' that is: of The probability of lirilurc for a value of U at random, that is the risk of failure R for an event of the type considered, will then be: 00 R= I 0 45 fo ( u).pT ( U).dU . . . (7) ISr3716-1978 dR a- ---------_-a-- Pr (ll) * = P,(Uj*fdU'~ -J dU R-7P Fro. \AREA Au OT (U) fo(U)dU R I #hadedarea OF THE RISK OF FAILURE OF A INSULATION 4 EVALUATION PIECE OF This expression shows the general principle of the method by which the probability of failure may be assessed. It assumes that_& ( U) and PT ( U) are uncorrelated. Non! -In principle formula (7) applies to a single-phase piece of insulation only. If several pieces of equipment, connected in parallel on the nave phase, are subjected to the same overvoltage then it may be assumed that the overall risk is equal to that of a single piece of equipment multiplied by the number of piccu in parallel. Thia is valid if we take into consideration the fact that the risk of failure acceptable for subrtation insulation ia usualiy very low. Sometimes it is necessary to evaluate the risk of failure of at least one phase of a three-phase section of the system following a switching operation ( for example a closing operation ). This risk may be obtained by multiplying by 3 the risk evaluated according to formula (7) if the probabil&y density f. ( U) of the overvoltages may be assumed to be equal on all three phases. _c An alternative method is to establish the overvoltage probability density f.(LJ )by considering only the highest value of the overvoltages caused on the three phases by a switching operation. Then the risk of failure ia evaluated by making use of formula (7). 46 The former approach gives risk values higher than the actual ones; the latter lower. Obviously the two approaches give results differing by let43 than 1 to 3. The mathematical overvoltage in formula ( the following assumptions a) Peaks other than are disregarded. model chosen for defining the severity of an 7 ) is based upon a few simplifications. In fact are made : the highest one in the wave-shape of overvoltages b) The wave-shape of the highest peak is assumed to be equal to that of the standard switching or lightning impulse. c) The highest overvoltage peaks are assumed to be all of the same polarity; to be on the safe side the more severe polarity will be wed. As regards switching overvoltages, which are the overvoltages of predominant importance in insulation design of EHV systems, assumption (a) is such as to give a calculated risk of failure lower than the actual risk. Assumptions (b) and (c) result in a calculated risk higher than the actual one, since the standard wave-shapes are so chosen as to establish the lowest withstand of apparatus (see 7.2 of IS : 2165-1977* ). In general, considering the opposite effects of the assumptions made, the risk of failure calculated by means of formula (7) gives risk values greater by about O-5 to 1 decades ( 5 to 10 times ) than the actual values. Normally formula (7) is, therefore, conservative. As said above, formula (7) may be applied for all the specific types of events significant in insulation design. Furthermore, it is clear that the accuracy in risk failure greatly depends on the accuracy in the overvoltages and of the impulse discharge probability accuracy of thesa is seldom satisfactory, the accuracy of failure can be correspondingly poor. the calculation of the determination of the of insulation. Since of the calculated risk However, the risk of failure has a precise physical meaning ( contrary to the conventional safety factor ). By making use of statistical methods it is therefore, possible to coordinate the security levels of the various parts of system according to the consequences of a fault. Furthermore, it is possible to carry out sensitivity analysis ( for example, the effect of a change in the overvoltage severity or insulation withstand capability on the probability of faults ). Statistical methods do, therefore, enable the engineer to take a decision on a rational basis. According to the statistical method, insulation is selected in such a way as to obtain a calculated probability of failure lower than, or qua1 *Insulationco-e&nation ( suond mmon ). 47 IS : 3716- 1978 to, a predetermined value that characterizes the required safety, level. Referring to Fig. 4 a change in the insulation level shifts the curve reprcscnting the discharge probability of the insulation PT ( U) along the U axis with a consequent modification of the shaded area A which represcnts the probability of failure R for U at random. The statistical approach may require successive series of tentative designs and evahlations of risk, until a design that satisfies the risk prerequisites is found. Formula (7) may also be applied to determine the probability of failure of an insulation protected by spark gaps or surge arresters, if PT ( U) is taken as the discharge probability of the insulation in presence of the protective device. If the time-to-discharge of the protective device can be_ considered always shorter than that of the insulation to be protected an equally valid and simpler method is to use formula (7) and to take&( U) as the overvoltage probability density modified by the protective device ( see Appendix C ). The use of digital computers makes it easy to evaluate the risk of failure, and therefore the insulation design, once the overvoltage distribution and the discharge probability curves of insulation are known. 9.29 Simprified Statisticul A&hod - Sensitivity analyses and ready evaluations of the risk of failure may be made on the basis of simplified statistical methods in which the calculations are performed once andfor all hy making some generally acceptable assumptions concerning the mathematical laws by which the actual distributions of the overvoltages and of the discharge probability of insulation are represented for example by assuming them to be Gaussian with known standard deviations.~~With these assumptions the complete distribution of overvoltage and the discharge probability of insulation may be defined by one point only corresponding to a given reference probability and called in IS : 2165-1977* `statistical overvoltage' (~6.219) and `statistical impulse withstand voltage' ( see 222) respectively. The risk of failure may be correlated with the margin between these two values, so that the approach becomes rather similar to that in the conventional method. `_Q Figure 5 gives a graphical expl -on of the method. Figure 5A shows frequency distributions of wervo "$" `and insulation strength, where the statistical overvoltage is indicated by' ?? s and the statistical withstand In Fig.`SB, the overvoitage/distribution and the electric voltage by V;. strength distribution are superimposed for three values l-0, 1.2 and 1.4 of the statistical safety factor ( y ) relating U, and Ur. The correlation between statistical safety/factor and risk of failure is given in Fig. 5C. * EB : 3716 - 1376 3 IU) 100% -"--,`-' l=d!!d ---w 90% UW U 58 Three attempts at determining the risk of failure [ area A) for statistical safety factors.. UW - l-P, o.,, IUli SA US = statistical overvoltage R placed on the probability density curve [ the shadad area (2%) reprarentsthe rofermce probability]. UW = statistical withstand voltage Irced on the discharge probaEIlity curve (90% represents the reference probability ). Iy=l*O* us i-2 and I.4 SC Relation batwaen thr rtatlsticai safety fmor ( y ) mnd the riJr of failure ` R ' ( nnrrurrd by +ru A). ~IYPLIPIBD STATISTICAL MTitOD FIQ. 5 49 ts I 3716 - 1978 The reference probability of the overvoltages is chosen in this standard equal to 2 percent. The reasons of this choice are discussed later in this clause. As regards the reference probability of the withstand voltage the 90 percent value was chosen in IS : 2165-1977* for the reasons given in 4.3 of this standard. As regards switching surges, Fig. 6, 7 and 8 illustrate the rehitionship between risk of failure and the statistical safety margin for air insulation in different cases. The discharge probability curve of insulation was assumed to be Gaussian as stated in 4.3 with k ( formula 5 in 4.3 ) equal to 1 and er ( formula 2 in 4.3) equal to 6 percent, 8 percent and 10 percent ( see Fig. 6, 7 and 8 respectively ). If t were to be taken as differing from 1, the statistical safety factor given for k = 1 would have to be multiplied I by + Figure 6 is applicable to laboratory conditions, while Fig. 7 is generally applicable to service conditions. Figure 8 may be used for particularly severe conditions (high values of an in formula 2 >. In all the three cases the overvoltage distributions were assumed to be Gaussian truncated at three and four times the standard deviation aBr or not truncated and with standard deviation Usequal to 10 percent, 15 percent and 20 percent. Figures 6 to 8 give the average correlation between statistical safety factor and risk of failure as well as the upper and lower enveIopcs of the correlations obtained when considering the nine overvoltage distributions resulting from all the possible.combinations of values of standard deviation and upper truncation point. The choice of a Gaussian distribution to define the overvoltage severity does not mean that other distributions ( for example extreme value distribution) may not give better approximations, but that Gaussian distributions match actual distributions reanonably well over the range of interest. The correlation between the statistical safety factor and risk of failure appears to be only slightly affected by changes in the shape of the ovcrThis is due to the fact that the 2 percent value voltage distribution. chosen as a reference probability of the overvoltages falls in that part of the overvoltage distribution which gives the major contribution to the risk of If, on the contrary, a much lower f:$lure$n the range of risk considered. or hi her value were chosen, the influence of the shape of the overvoltage distri % ution would have been very pronounced. Figures 6 to 8'give the risk of failure of a piece of single-phase equipment ( for example a post insulator ). If the risk of failure of several pieces of equipment is required reference may be made to the Note under 9.2.2. *Insulationco-ordination ( second revisiin ). 50 IS : 3716 - 1978 R 1 8 P 10-l 6 I m-m-a-. I I I 1 Dlrttlbutlon truncated at 3 `* Dlstribullon truncated at ' " Dlrtrlbutlon not truncated I I I I I %A \ I I Upper mvrlopo o;l 08 occ 1 1,l 14 14 114 1,s 1,4 1,f Standard deviation of ov~rvoltagc distribution w, Standard deviation of inrulation q - 6 percent 10, 15 and 20 percent FIO. 6 CORRELATIONS BETWEEN RISKOF FAILURE (R) AND STATISTICAL SAFETYFACTOR (y )POR VARIOUS Swx~cxmo SURQE DISTRIBUTIONS 51 Stamiard deviation of insulation a~ - 8 percent FIG. 7 CORRELATIONS BETWEEN RISK OF FAILURE ( R ) AND STATISTICAL SAFETY FACTOR ( y ) FOR v~Rr0u.s !h'lTCliINt3 !bROa, hTJUBUTIONS For example, if the number of single-phase pieca of equipment at a line entrance is equal to 21 ( seven on each phase ) and the risk of &Sue of each for a &ret-phase reclo&g is IV, then the risk of firilun fbr the whole line entrance will be clou to 21 .lW. ts Extension of this method to ove&ead lizua ir p&ble but problems particularly for long lines with which the present guide iETn not attempt to deal. 52 I6 I 3716 - 1918 Standrrd deviation of insulation UT = 10percent Fm. 8 CORRELATIONS BETWUIU RISK OF FAILURE ( R ) AND STATISTICAL SAFETY FACTOR ( y ) FOR VAR~US SWITCHING SURGE DLWRIBUTIONS As regards lightning surges, analogous correlations between statistical safety factor and risk of faihrre are given in Fig. 9 to 11 for comparison purposes always for air insulation. Assumptions similar to those of the previous paragraphs were made as regards the discharge probability curve of insulation to lightning surges. The lightning overvoltage distributions were assumed to be Gaussian d 60 and not truncated, with standard deviations equal to 40 percent percent. It is thought that such di&ributions approximate actual ligI?tning overvoltage distributions quite well around the 2 percent value. 53 I8 83716 - lmb Figures 9, 10 and 11 give the correlation between statistical safety factor and risk of failure for both overvoltage distributions and for standard deviations of insulation equal to 3 percent, 5 percent and 7 percent respectively. NOTII- The correlations between statistical safety factor and risk of failure given above apply to relf-restoring insulation only. However, they may be considered acceptable for the entire equipment in most cases for the following reasons: The tats da&bed in 7.4 and 73 of IS : 2165-1977. are intended to ascertain the impulse insulating ability of self-restoring equipment and combined equipment respectively. Such tats do not allow any discharge on the non-self-restoring parts of the apparatus. Consequently, when both the self-restoring pieces of insulation of the same apparatus are designed on the basis of the same risk of failing the test, the non-self-restoring parts will have an inherently,. lower discharge probability than the self-restoring parts in respect of overvoltages of the same amplitudes as those of the impulses applied during the test. Consequently, it may be stated that risk failure of the non-self-restoring parts of self-restoring equipment ( tested according to 7.4 of IS : 2165-1977' ) is lower than tbat ofthe self-restoring parts if the major contribution to the overall risk of fiilure .ir given by overvoltages in the range of U56y0 * 20. For combined insulation equipment the non-self-restoring parts will be designed for a low risk of failure at test voltage. In the cases where the major contribution to the failure risk ( calculated on the basis of the statistical withrtand voltage ) is given by overvoltages around the test voltage the evaluation of the risk of failure may be carried out as for self-restoring insulation. The foregoing considerations naturally presuppose that wave-chopping by a self-restoring piece of insulation does not cause serious stresses in the non-sclfrestoring insulation of equipment and does not cause ageing of insulation. 9.3 Block Diagram of the Insqlation Design and Coordination of an Electric Installation - Most predictions or analyses of system overvoltage levels assume that a piece of equipment ( for example, a circuit breaker ) will operate as designed. In other cases, an arbitrary limit may be placed on the `credible' severity of surges, as is often done in the case of lightning. It is obvious that surge levels based on such assumptions will sometimes be exceeded. Whether or not it is necessary to take this into account, depends largely on the consequence of failure resulting from such abnormally high overvoltages. For example, the consequences of transformer or reactor failure are so serious that there insulation co-ordination must usually provide for even extreme contingencies. This is achieved by applying surge *arresters at their terminals. There are other types of equipment such as post insulators, disconnecting switches, etc, where the consequence of failure is not so serious as to warrant application of surge arresters. In developing a squence of insulation design and co-ordination of an electrical system in the form of a block diagram, it is convenient to l InrJation co-ordination( sacodrmirion ). IS:37169,I978 0.5 0.6 0.7 (I-80.9 1-O 1.1 1.2 1.3 y Standard Standard deviation deviation of insulation of overvoltage UT 3 percent o8 = 40 aqd 60 percent distribution Fro. 9 CORRELATIONS BETWEEN RISK OF FAILURE (R) AND STATISTICAL SAPETY FACTOR (y) FOR VARIOUS LIGHTNING SURGE DISTRIBUTIONS differentiate between a Gas:: I ( surge arrester protection) and Case II ( no surge arrester or remote surge arrester protection). A block diagram illustrating the method to be adopted is shown in Fig. 12. The first step in insulation co-ordination ( Block 5 ), common to all types of equipment, has the purpose of ensuring the eqtlipment ability of withstanding poweti-frequency voltage under normal conditions and under tempdrary overvoltages. The system engineer will specie an equivalent pollution severity test level for insulation susceptible to contamination, however, no special specifications will be given for insulation susceptible to ageing ( see 9.1). If the expected phase-to-earth temporary overvoltages ( Block 3 ) are more severe than the overvoltages taken into consideration in the relevant cquipmznt specification in specifying the power-frequency tests discussed 55 ISt3716-1978 R O-5 O-6 $7 0.j 69 l-0 l-1 I.2 1.3'f Stmdard deviation of imuhthn a-r - 5 percent Standard deviation of overvoltage distribution e, - 40 md 60 percent Fxa. 10 CORREWTION~ BSTWG~NBUK OF FAIL( R ) AND STATISTICALSAFETY FACTOR ( y ) FOR VARIOVS LIOHTMNO thltOE hWUlWTIONS in 3.6 of IS :2?65-1977*, it will be necessary to specify different voltage levels or durations of the test or to ado t suitable means or operational procedures to reduce temporary overvo P tages in the system (feedback from Block 5 to Block 2 dashed line). Insulation dosign as regards operating voltages and temporary over-voltages leads to a certain withstand of the equipment to both switching For instance, if a given withstand salinity is and lightning impulses. required for a post insulator, a minimum distance in air is obtained which varies according to the post insulator type (see Appendix E ). The equipment will, therefore, exhibit a certain withstand to switching surges due to requirements imposed by the operating voltages and temporary overvoltages. These influences are indicated in Fig. 12 by dotted lines ( for example, from Block 5 to Block 7 and Block 8 ). *Insulation co-ordination (second reuisioh ). 56 O-5 O-6 0;7 O--80;9 FIG. 11 1;l 112 1.3 I' Standard deviation of insulation OT= 7 pacat Standard deviation of ovcrvoltagc distribution ea - 40 and 60 percent CORRELATIONS BETWEEN RISKOF FAILURE(R) AND STATISTICAL SAFETY FACTOR (y) POR VARIOUS LIOHTNINO SURGEDISTRIBUTIONS Then it is necessary to consider Case I and Case II separately. Examples of selection of the rated switching and lightning impulse withstand voltage are given in Appendix D. As regards apparatus of the first type ( Case I ), choice of the rated switching and lightning impulse withstand voltages is usually made as follows: Choose the rated voltage of the surge arresters on the basis of the temporary overvoltages (Block 6 ); thus the protective levels of the arresters under switching and lightning impulses will also be determined at least within certain limits. a> b) Choose the rated switching and lightning impulse withstand voltages of the apparatus on the basis of conventional safety factors dictated by experience ( SCI 5.4-l and 5.5.1). 57 ISr3716-1978 C) Adopt suitable means in system design or suggest suitable operating procedures to reduce temporary overvoltage if an economic incentive exists to reduce insulation levels ( start again from Block 2 ). This procedure of selecting the rated switching and lightning impulse withstand voltage of the apparatus disregards the severity of the actual switching and lightning overvoltages by which the apparatus connected in parallel with the surge arresters may be stressed since it is based on the protective level of the surge arresters only. The insulation levels of the apparatus according to the following steps: of Case II are usually chosen a) Choose the rated switching-impulse withstand voltage of the apparatus on the basis of an acceptable risk which may be estimated directly on the basis of the expected distributions of the overvoltages and of the discharge voltages (statistical method), or by means of the correlations given in 9.2.3 between the risk of falure and the statistical safety factor ( simplified statistical method). Actually these correlations apply to selfrestoring insulation only, but norinally they may be considered acceptable for the entire equipment ( SM Note under 9.2.3). NOTE -Protection against switching surges afforded by surge arresters installed close to equipment in Case I and gaps installed at the line-entrance may often be discounted as regards Case II apparatus for the following reasons: i) Most of the types of equipment belonging to Case II ( especially the lineentrance apparatus) may at times be isolated from the surge arresters installed in the station to protect Case I apparatus; ii) With the present technology of surge arresters, the protective level against switching impulses is often greater than, or equal to the highut awitching overvoltage which may occur with correct bebaviour of system apparatus. Insulation must therefore be designed to withstand these overvoltage; and iii) Spark gaps may not provide a substantial degree of protection against This point is switching surges if undesired sparkovers are to be avoided. discussed in Appendix C. In Fig. 12 the block between indicated by dashed lines. Block 4 and Block 8 is therefore b) Adopt suitable means in system design to reduce switching overvoltages, if this is possible and if an economic incentive exists in reducing the rated switching impulse withstand voltage of apparatus. No economic incentive may exist in reducing insulation : for instance if the withstand to normal operating voltages and temporary overvoltages calls for ` higher insulation ' than switching surges ( dotted line between Block 5 and Block 8 ). 58 1 1 Oparatlng voltage under normal condltlonr 0 J I f 1 0 Charactariaticr of thr ryatom Calculation of temporary ovrrvoltagrr 0 lntrodktion of moans to reduce trmporary or awitchlng ovarvoltrgra 0 overvoltagaa J ir----L---- Salecttion of surge arresters rated-voltage I 0 1 ' j 1 ! 0 Insulation design aa regarda operating voltages and temporary ovarvoltagsa l I 1 I L ----- PoeaMe protscuvo I effect of ourgo arroatora and spark gape ----I I I t , L 1 I ........ ... ...... ...*............. .... .. .(I..,..........,.......,*........*... ..: ! 1 i :a.........,......... *.... . .. ..a,. i 1 T Selection of ratad lightnlng and awltching withstand impulao voltage of Case I apparatus ! 1 0 lntroductioc of means to raduce llghtning ovorvoltagra ----_---------------l I I 4 I Selectlon of rated and switching withstand Jmpular voltage of case II apparatus fLine insulation level, number of earth wires spark gape at the line-entrance 4 69 0 I i Lightning overvoltage characteristics 4 Calculatton of lightning overvoltage0 b 4----Selection df rated lightning withstand impulse voltage of case II apparatus 0 Fm. 12 BL~~K-D~~~AY OF INSULATION CO-ORDINATION AND DESION As in the Original Standard, this Page is Intentionally Left Blank IS : 3716 - 1978 4 Verify that the rated lightning impulse withstand voltage, corresponding ( in Table 3 of IS : 2165-1977* ) to the rated switching impulse withstand voltage determined above [ Items (a) and (b) 1, guarantees a satisfactory performance of the apparatus under This'should be done on the basis of the lightning overvoltages. expected distribution of the lightning overvoltages and of the discharge voltages or by means of the correlation given in 9.2.3 but for the sake of simplicity it is often done on a conventional basis ( see Appendix D ). Consider that only the highest value of rated lightning impulse withstand voltage of each line should be used for apparatus not effectively protected by a surge arrester ( see 6.4 of IS : 2165-1977* ). Provide means to reduce the amplitude of lightning surgest (feedback from Block 12 to 10 ) or choose a rated lightning impulse withstand voltage higher than the one determined on the basis of Table 3 of IS : 2165-1977* if too high a risk of failure to lightning overvoltages results from Item (3). In the latter case the value of the rated lightning impulse voltage shall be selected from the series ,in 6.1(b) of IS : 2165-1977*. 4 APPENDIX ( Clauses 351.6, SURGE A-l. TRANSFERENCE DERIVATION THROUGH A TRANSFORMERS 7.2.2.1, 7.2.2.2 and 7.2.2.3) OF EXPRESSIONS A-l.1 Initial Capacitive Voltage Spike -- During the initial period of about one microsecond under the conditions of a lightning surge, the transformer may be approximately represented as shown in Fig. 13A as a capacitance voltage divider of ratio s where s < 1. If Ct is the sum of the capacitances of the higher-voltage and lower-voltage arms of this divider, the initial transference may be simulated as shown in Fig. 13B, 13C and 13D by a series circuit comprising a source U. = sU,, a capacitance Ct and the capacitance or resistance of the external lower-voltage system. Vi is the surge voltage on the higher-voltage side during the initial period. The source voltage U, is the open-circuit transferred voltage. l Insulationeo-ordination( second revision) . jConsider possible changer in line design such as tower footing and shielding wires, install surge arresters other than those intended to protect Case I apparatus, make use of protective spark gapr. The actual degree of protection provided by spark gaps is dbewud in AppendixC. 61 13A frrnrformar Voltage 8s a Gpecltivr Divider 130 Equlvdont Circuit of Cep8dtwue Trw4emksion R I3C External System having Gpacitrnce Ce 130 Ext;zzRrn lmviy Fm. 13 INITUL CAPMXTNE VOLTAOE Srmx If the external system may be represented by a capachnce C. as in Fig. 13C the equivalent circuit is a voltage divider having tbe ratio: ct c; + c, . .. (8) If during the initial period, the impedance of the external system is the surge impedance of a cable or the resistance of the load, this system may be represented by a resistance R as shown in Fig. 13D. Typical values are from 10 ohms to a few hundred ohms. The transf'erred surge voltage is then dependent on the steepness as'well as on the amplitude of the surge. For high values of R the initial voltage is approximately & and for low ' values of R it is given appr ximately by Us = s S R Ct where S is the maximum steepness of the sz rge. The above ex$essioru do not take into account the effect of superposition of the surge voltage on tbe power-fiqucncy voltage. 62 Isr32w-Em m may be made for the power-frequency voltage by substituting ti Ur the actnal peak voltage Us and by introducing the fstor #. For a star/delta or delta/star connected trarqformer the value of fi is typically about 1-15. Far a star/star or ddta/ddta connected transformer the due of p b typicauy ablxit I-05. However, slightly higher values than For switching surges the value of p may be theW?maybeeWXmWWL taken as unity. The amplitude U, will be limited to the front-ofwave sparkover value of the surge arrester or spark gap on the higher-voltage rideofthetrausfbr~r[truIS:3070(PartI)-1974*]. side is given by: The amplitude of the initial voltage gspike ' on the lower-voltage u, = JWR . .. ... (9) i?or a transfixmer without external connections to the lower-voltage term&u& the value of fxtor J may range from zero to at least 04, The value of s may be measured depending on the winding arrangement. in a low-voltage impulse response test ( for example, with a recurrentsurge osoiliogrnph ). Values of Ct generally lie in the range lW* to lO_' farad. the mmn~turcr Z$xra- The valua of 8 & Ct 8~ di&ult to calculate bra new transformer MC may only be expected to give a rough estimate without guarantee. The vab of U, should be compared with the appropriate test voltage of Table 1, List 1 or 2, of IS : 2165-1977t. impulse The amplitude of the transferred surge may be reduced by: sparkover a) Using a surge amstcr with a lower front-of-wave voltage OIIthe higher-voltage side, between each phase and earth on the lowerside between e&cl b) A&Gig capacitance voltage side, and c) Alding a surge arrester on the lower-voltage phase and earth. For mm&cal example see Aa. A-l f WeIy W~E&WWI V&age - The transference of emf by inductive coupllug between windings in a 3-phase tramliumer or bank of trar&rmers may be evaluated fbr any winding connections by considerii the surge voltage as . a single-phase alternating voltage. The e&ct of delta whdimgs on the zero-phase-sequence of the siuglqbam voltage should be taken into account. -4 . . -tadC?Id8gerrmat systems: Part I Non-lkar l&kg-r--vmuiation~ (smadrmisirr). 6.3 compouent resistor type IS-: 3716- 1978 Figure 14 shows the results for eight different connections transformer, assuming the system voltage ratio is unity. of the As in the analogous power-frequency considerations the transferred voltages at the terminals are determined by the emf's and by voltage division between the internal impedance of the transformer and the external circuit impedance. The efm's may be assumed to have the same waveshape as the surge on the higher-voltage system, if the effects of internal oscillations in the windings are neglected. The response of the lower-voltage system to these emf's is usually in the form of a voltage of similar shape to the surge with a superposed oscillation. The amplitude of the inductively-transferred voltage depends also on the voltage ratio and 3-phase connections of the transformer and on the relative impedances of the lower-voltage system and the transformer. The voltage on the lower-voltage side of the transformer is given by: G-P where 4 = is a response factor of the lower-voltage circuit to transferred surge emf, r = is a factor depending on the transformer connections Fig. 14), Ua - is the peak voltage to earth on the higher-voltage side, X = is the system voltage ( phase-to-phase) ratio of transformer. the (see and the 9r G/N ... ... (19) The value of q depends on the waveshape of the surge and on the electrical parameters of the lower-voltage circuit. For lightning surges on a transfermer having Category 1 equipment without appreciable load connected to the lower-voltage aide, the val,ue of q is generally not greater than about 1.3 although this `value, may be exceeded. For switching surges on a similar system without appreciable load, the value of q is not greater than about 1.8. Generally lower values of q apply if an appreciable, load is connected due to voltage division between the load impedance and the leakage inductance of the transformer ( su#Note ). For Category 2 equipment, voltage division takes place betweenthe leakage inductance of the transformer and the subtransient inductance of the generator, and if these are about equal, q has the value of about O-9 for lightning and switching surges. Values oft for a surge on one phase only ( for example, a light&g surge) and for equal surges of opposite polarity on two phases ( one .fYPe of switching surge) are shown in Fig. 14 for eight different 3-phase connections of the transformer. 64 As in the Original Standard, this Page is Intentionally Left Blank I8 : 3716- 1978 The calculated value of Us is an estimate of the longer-duration transferred voltage, which in practice includes longer-term effects of capacitive transference and transferred voltages corresponding to internal oscillations within the windings. Its amplitude will be limited by the protective level of the surge arrester or protective spark gap. In the case of the former this will be the higher of the standard lightning impulse sparkovcr value and the residual voltage value for the lightning surges [ JC~IS : 3070 ( Part I )-1974* 1. For switching surges, except when the transformer is connected to a highly-inductive load, such as an induction motor, over-voltages on the higher-voltage side may be assumed not to exceed per-unit overvoltage of 3 (see Note below ). .I The value of Us should be compared with the peak values of the appropriate power-frequency test-voltage of Table 1 of IS : 2 165-1977t. It may be found necessary to reduce the value of the lightning or switchAdding extra ing surge on the higher-voltage side of the transformer. capacitance to the lower-voltage side has little effect upon the amplitude of the inductively-transferred voltage but it may be desirable to consider the addition of surge arrester. NOTE- When the circuit is switched off on the higher-voltage side of a transformer which may be loaded on the lower-voltage side by reactors or other inductive load, then dangerous overvoltages may be attained under the most unfavourable conditiom of operation but, in general, Us does not exceed the peak value of the power-frequency test voltage since Q is less than 1-O on account of voltage division between the trans. former and the load inductance. A-2. NUMERICAL A-2.1 Example A - EXAMPLES Category 1 Installation Transformer Lightning impulse withstand voltage Impulse test voltage 145 kV star/l2 kV delta 145 kV side 12 kV side = 550 kV = 60 kV (assuming List 1 of Table :165-%37;;j Power-frequency Rated voltage test voltage 12 kV side =28kV = 120 kV Surge arrester on higher-voltage *Lightning arresters for alternating lightning arrestera (jrst revision ). tInstrIation co-ordination current side of the transformer: systems: Part I Non-linear resistor ( secoad repision ). 67 Front-of-wave sparkover vdtrgs 1*2/50 impulse sparkover @MI residual voltage a) Liyning swgeon onr#hasb ==463kV =4OOkV I Initial voltage `spike' for transfbrmer disconnected from loadz U, - roUp - 04 x I.15 x 463 = 213 kV Assuming a ratio of I.25 between the impulse withstand test voltage and the service overvoltage, cable connections f&n ct that is, --&w 48 kV, would require the of: the transfbrmer to have a capacitance ct+c, hence Cm> 3'44 Ct %i3-O'23 48' Having obtained from the transfixmer manufacturer a value of farad, the cable capacitance per phase should be ;?t +st 942--x 1W farad. Ct = IV If the load is connected then this will reduce further the p&cVd$age on the lower-voltage rid&. Due to inductive transferarce : u, = Pe &IN 1.15 x 1.3 x 0.577 x 400 c9 28_5 kV 12.1 where W5 kV gives a ratio of 1*4 with the 39.6 kV peak test voltage (28 4 p) so that the insulation to Tabk 1 List 1 of IS : 2165-1977* shall be satirfhc~. b) Wcking~~esofoppofitrjo&@yon&uo~hass Amuming a PA. p= l*O,qswitching ovwal&ge ori two phases of 2.5, and 2 1*8,andr= l-15, then: +K 1-O x.1.8 x 1.15 x 145 x ,/2`- x 2-5 U _ ? d/sx 12'1 =5@5kV ls : 3714 - 1978 In this case the voltage peak exceeds the peak of the test voltage of 39.6 kV for Table 1 of IS : 2165-19778. To allow for this either insulqtion for List 2 of the Table 1 of IS : 2165.1977* shall be used ()hat is, ha&g test voltages of 75 kV lightning impulse withstand, and 28 kV rrns/39*6 peak) or switching surges between phases on the higher-voltage transmission system shall be limited, alternatively a surge arrester may be selected on the higher-voltage side of the transformer havitig a lower protective level or consideration may be given to fitting a surge arrester on the lower-voltage side. A-2.2 Example B-Category 2 Installation A-2.2.1 Typical &amlpIe for 220 kV Transformtrs Delta generator transformer 245 kV stat-124kV Lightningimp& withstand voltage Power-frequency test voltage Imp&e test voltage 245 kV side = 1'050 kV ( Ref. Table 2 of IS : 2165-1977* ) 24 kV side = 50 kV 24 kV side = 125 kV ( Ref. List 2 of Table 1 of IS : 2165-1977* ) transfornur: 198 kV 746 kV 649 kV Surgs arrester on higher-voltags sid6 of = Rated voltage = Front-of-wave sparkover voltage l-2/50 impulse sparkover and = residual voltage a) Li&?rir& SUf~6SOn-Otl6~hU~6 Initial voltage system: ` spike ' for transformer disconnected from 24 kV G#==&J, = O-22 x 1'15 x 746 = 189 kV ( maximum)' ( the value for s having been obtained from the transformer manutacturer ). This value would obviously be too great for insulation on the lowervoltage side. &tuning a ratio of I.25 between the impulse test voltage and the service overvoltage, that is a value. not exceeding 125/l-25 `= 100 kV In this case: _ would require the addition of external capacitance. ct ct = c; hence C, > O-885 Ct *Insulation co-ordination ( second mkion ). ' 100 = O-53 189 69 IS I 3716 - 1978 The value of Ct is obtained from the transformer manufacturer for a wave steepness S = 1 200 kV/ys. The dther methods listed in this Appendix for reducing the capacitive peak ( use of surge arresters ) may ah be considered. The overvoltage due to inductive transference is:. U, = pqr wfl 1.15 x 1.0 x 0.577 x 649 = 41.5 kv = 10.2 41 kV is less than the 70.7 kV ( 50 1/r ) peak tests voltage, and this gives a ratio of 1.7, which can be considered adequate. b) Switching surges of opposite polarities on two pfzases Assuming a p.u. switching overvoltage on two phases of 2, then the inductive transference is: u, = where p = 1.0, 4 = l-0, and 2 r = = 1.15 tir 43.5 kV is about equal to the inductive transference of lightning (41.5 kV). A-2.3 Example C Category 2 Installation I.0 x 1.0 x 1.15 x 245 x dflx 10.5 x VT 2 = 43.5 kV surge Ty@*calExamplr& 300 k V Transformers 300 Delta generator transformer 300 Lightning impulse withstand voltage 24 Power-frequency test voltage 24 Impulse test voltage kV star/24 kV kV side = 1050 kV kV side kV side = 5.0 kV 125 kV ( assuming List 2 ofTable 1 of IS : 2165-.1977* ) Surge arrester on higher-voltage side of transformer: Rated voltage Front-of-wave sparkover voltage l-2/50 impulse sparkover and residual voltage lInsulation co-ordination ( smnd d&m ). = 240 kV ==9OOkV = 785 kV 70 IS : 3716- 1Y-m a) Lightning surge on one phase: Initial voltage ` spike ' for system: transformer disconnected from 24 kV U,s = spup = O-2 x l-15 x 900 = 207 kV maximum, the value for s having been obtained from the transformer manufacturer. This would obviously be too much for the insulation on the lower-voltage side. Assuming a ratio of l-25 between the impulse-withstand test voltage and the service overvoltage that is a value not exceeding 125/l-25 - 100 kV, would require the addition of external capacitance. In this case Hence C, ) Ct c, + c, < , _!!! 207 = O-48 1.08 Ct The value of Ct is obtained from the transformer manufacturer for a steepness $ - 1 200 kV per microsecond [see IS : 3070 ( Part I )-1974* 1. Alternatively the other methods listed in this appendix considered. Uue to inductive transference: G = kf pi -&IN 1.0 1;.50*577 x 785, 41'4 kv 1.15 x could be 41'4.kV is less than 70.7 kV, ( 50 42) test voltage and this gives a ratio of 1.7 which may be considered adequate. b) Switch& surges of opposite polarity on two #iases: Assuming a p.u. switching overvoltage on two phases of 2, then the inductive transference is: u _ 1.0 x 1.0 x 1.15 x 300 x dT_x 2 2- 12*51/5 -45kV where 2 = 1.15 tiF 45 kV is about equal to the inductive (41.4k.v). t- P = 1.0, 4 = l-0, and transference of lightning surge resistor type l Lightuing arresters for alternating lightning arresters (first reviiion). current rystetns : Part I Non-linear 71 APPENDIX (C'imssc 4-1.2) B B-l. GONFlDENGE LIMXTS OF THE TEST Tests may provide only more or less accurate estimates of the true values of the withstand strength of equipment. An increase in the accuracy may be obtained by an increase in the extent of the test. The exttnt of tests must, however, be limited for reasons of costs, the diminishing return of gain in the accuracy and possible destructive efktx on the equipment. For these reasons IS:2165-1977* prescribes three difkent test methods in 7.2,73 and 7.4 according to the type of equipment. In 7.3 ( a ) dealiig with the 50 percent disruptive discharge test, note states that ` There are a number of procedures available, and any of these may be used provided that the accuracy of the determination is within one half of the standard deviation with a confidence level of 95 percent ' . It can be said that there should be a 95 percent probability that the 50 percent discharge voltage of the equipment at the time of the test be within the boundaries given by the value estimated from the test plus and minus one half of the standard deviation. the A test procedure fulfilling this requirement is the ` up and down ' test with 30 shots. The accuracy of the 15 impulse withstand test is considerably less. The 95 percent confidence limits for the probability of discharge are in this test: For 0 ,, ,, ,, ,, ,, 1 2 3 4 5 discharges >. ,P ,P ,P ,Y approximately 3% 3, ,P ,) 9, 0 0 0.015 0.045 088 0.12 to 0,213 0.32 0.40 0.48 0.54 0.61 ,, f) ,, ,, ,, Based on 95 percent confidence limits, it may be seen that irom a test with 15 impulses only, it is not possible to conclude that the discharge *Insulation co-ordination ( second rmisim ) . 72 Bt37l6-1976 probability is lets than @lo, or that the probability of withstand is higher ~~~crrarfforodipehrse,~whenrhenumbgofdirchorgesexceadr 4,thctcstisaigdcantta tut dthe hypothuia that the probability of wiwzmdiagiskss thano90. B-Z. EXAMINATION OF DIplpIsRBNT METHODS OF TEST Accept@ the above basic IimMions of tats with a small number of impulses,the Mlowing exposesthe validity of differentmethodsof discharge te&ng and tbc balance between the risksto the manufacturerand the purchaser bearing in mind the tical necessityfor a limited number of impulse8and, for .the mantlEzcuter, to design his products to have an economicallyacceptabk risk of&hare on tests. For the sake of simplicity it is assumed in this apptmdix that the dischargep&ability Pt (U) df the diffcnnt piecesc&equipment in a given populati~ ( same type of equipment on which an impulse test U to be canicdaut)fdkwsaGaussian law with a constant standard deviation ut cquaI for all the pieces ofequipment. In this caseone parameter ( X given in pa ofat, s88 Fig. 15) is ru&knt to determine the dcviion of the inSulatio~strength of one piece of equipment `2 at the time t ( VW ) from thcspec&dval~ (Uaw). -1 -0.5 0 05 1 XI UlIo- URW 01 K IN p.u.OF Qt Fra. 15 DEFINITION OF THE INSULATION `STRENOTH OF A PIECE08 EQUIPMENT AT THB TIME 1 AS FUNCTION OF THB PARA-R X 73 XS: 3716- 1978 The 90 percent withstand strength of a piece of equipment in a population varies from one specimen to another. Fig. 16 shows how this may be described statistically in terms of K. The value of oP is very small for those types of equipment ( for example, disconnectors ) which may be considered essentially air insulating structures, since tolerances in dimensions are always very small. To ensure the repeatability of tests, ambient and insulation conditions should be kept as constant as possible during test ( or correction factors should be used ) and standardized testing techniques should be adopted. In principle, therefore, the discharge probability of a given piece of insulation under test conditions should not be expected to change. In other words the Pi (U) ( ac Fig. 15 ) should be the same in different tests of the same type. However the value of the 90 percent withstand strength of insulation may show variations from the average value derived from several tests by the same methods carried out in the same laboratory at different times or in different laboratories, due to differences in the ambient and insulation conditions or in the test circuits. How the laboratory inaccuracy can be described statistically considering the average value of the insulation strength of one specimen as the ` true value' is shown in (b) in Fig. 16. Assuming that the distributions in (a) and (b) in Fig. 16 are Gaussian with known standard deviations and that the design value of Uo has been chosen by the manufacturer, the probability density of the deviation of the measured 90 percent withstand strength of the population of a piece of equipment may be calculated [ see (c) in Fig. 16 1. An ideal test should be such as to prevent equipment having, at the time of the test, either an insulation withstand lower than prescribed to pass the test or an insulation withstand equal to, or higher than prescribed to fail. Self-restoring insulation having, at the rated impulse withstand voltage applied during the test, a probability of withstand equal to, or higher than, the reference probability ( 90 percent ) should have a probability of passing the test equal to 1, while insulation having a probability of withstand lower than the reference probability should have no chance of passing the test. The probability of passing an ideal test of an apparatus, the insulation of which, during the test, differs by X from the prescribed value, is represent& in Fig. 17 as a function of X by the solid line. Actual test, however, depart from the ideal test and follow in the K) plane curves similar to the dashed line. Figure 17 shows the curves for the tests proposed in 7.2, 7.3 and 7.4 As far as the test of 7.3 is concerned an `up and of IS : 2165-1977*. do- ) tat based on 30 shots was taken into consideration. ( 4, l Indation co-ordination ( srcondf&ion ). 74 IS : 3716 - 1378 RATED IMPULSE WITHSTAND VOLTAGE 0 KD a) Distribution of the 90 percent withstand voltage of the apparatus of population p. b) Distribution of the di!erence between the 90 percent withstand voltage of a given apparatus measured m various laboratories and the actual one. c) Distribution of the 90 percent withstand voltage of any apparatus of population p in various laboratories. FIO. 16 FREQUENCY DENSITYOF THE MEANRED 90 PERCENT ' WITHSTANDSTRENGTH OF A POPULATION OF APPARATUS The probabjlity density of the deviation of the measured withstand strength [ see (c) `in Pig. 16 ] of a given population is represented in Fig; 18 by Curve 1. By multiplication of the values of this curve by the value of the probability of passing a given test procedure as a function of K ( see Fig. 17 ) Curve 2 is obtained, This curve represents the density distribution of the 75 l3 t 37lS - 1978 I 3' / as/ f 8 03, IDEAL TEST If 2,' / //' I ?/" . 4 -1 -04~[).6-O*& -0.2 0 0,2 0.4 0.6 O-8 1 K IN p.u, of pI 1 - Up and down test 30 impulses ( ?.!Z of IS : 216~1977') 2 = 1512 test ( 7.3 of IS: 2165-1977. ) 3 - 3/O test (7.4 ofIS: 2165-1977') *Insulation co-ordination ( sacoadfaGon ). Fm. 17 PROBABILITY OF AN EQUIPMENT TO PASS THE DIFFERENT TYPE OF TEST AS A FUNCZIONOF ITS INSULATINQ CHARACTZRISTICS R of the population of apparatus for the given test accepted deviation procedure. If the area limited by Curve 1 and its abscissa is taken equal to 1, Area A in Fig. 18 represents probability Rm of rejection of a good product ( manufacturer's risk ) while Area B represents the probability of rejection of a deficient product. As a matter of fact the intended value of insulation design W will be chosen by the manufacturer on the basis of the sum of Areas A and B, that is, on the probability Pt of failing the test. Area C represents the probability R. of acceptance of a deficient product ( customer's risk ). By repeated calculation of Rc and PIfor different values of 76 la: 3716h78 ACTUAL rA C _I 0 \0 K IN p-u- OF at FICL 18 PROBABILITY Pt OF TEST FAILURE ( AREAA + B ) CUSTOMRR'S MANUFACTURER'S RISK % (AREA A ) &SK & ( AREA c ) the intended value of the insulation strength ( IV), curves can be constructed which show the relation between the risk of the user to accept a deficient product and the manufacturer's probability Ps of failing the test. Figure 19 shows such curves on the assumption deviations up, 01 and at are those stated in the figures. and Figure 20 gives the correlation between that standard risk ) RC ( customer's risk ) for the cases considered in Fig. 19. Rm ( manufacturer's It is to be noticed that if ea and ur are zero ( homogeneous population and no laboratory inaccuracy ) one of the two values Rm and R,, is zero. In other words there is onIy a risk either for the user or for the manufacturer. This risk as well as the value of Pr may be obtained directly from Fig. 17. 77 1!3:3716-1978 1 = 50percent 2 = 15/2 test 3 3/O tat discharge test FIO, 19 CU~OM~R'S RISK Rc AS A FWNCTION OF PROBABILITY P! m FAIL THE TEST 78 IS : 3716 - 1378 1 P 50 percent 2s 3 ~~~~ 20 15/2iest 3/o test discharge test CUSTOMER'S R&K R"R;~K~iF~~~~~~ QP MANUFACTURBR'~ m 79 I6 t 3716 - 1976 .APi'ENDIX C ( ckuse~ 5.4.1, 5.5.3 ilnd 9.22 ) C-l. For a given waveshape of the applied impulse, let us cali: the disruptive discharge probabilities of the a) pi (W-and 2 (2 msulatmn an of e spark gap as a fundion of the crest value U of the impulse*. b) Ptp (v) the probability that the insulation may discharge before the protective gap sparkovcr as a fbnction of the crest value U of the nnpulse. The discharge probability curves of insulation Pi (U) and of the spark gap pD (u) connected in parallel are expressed byz Pi(v> ic Pi(u)[ 1 pD (v) 3 + pi (v) PI (u) pip iv) . ..(Sl) PD(U)PPD(U).[l-Pi(U)l+PD(U)Pi Nom -On the asmm&~ that B spark gap Slow a Gauniam law, *I%im* PI, (V) ir cxplv=d hp_ fv)~i--ab~v)~-~*~~~ mzq MdlhF . . . (13) WhUC 2-,(u) T`(u) utP (v) ~thc5opcrcalt-cnlaeoftbctimc-~koverofthe~gap,ur fam&ndtbcnatvalueUofthc8pplialimpulse; irtlE5opCrcentvahleoftbetim. argeoftheinmh&qa8a fanctionoftheuutvahaeUofdreappliimp&q is the standscd deviation of the time-torpartovcr of the rpark gap, 081 (v) ~8functioooftheaat~eUofthe8pplicdimp~ arKI is the standard deviation of the tim~tmiischarge of the insuIation, a8 a function of the crest v&m of the applied impulse. If, for a particular combination ofthe insulation to be protected and of the spark gap, there is a negligible probability that the tin&-to-&+ charge of equipment may be lower than the time-to-sparkover of the gap in +he entire range 0 < U < Z&W, Pip. (U) becomes zero and fonnuh (11 may be written: &(V)ePi(U) ~l-pD(u)~ . .. (14) The formulae given below are gwxmlly valid for two piecu of insulation in paraW. 80 IA the case ofcombinationer dim&h ki&weshallconsidertbe~oftbeprot~vegapasbeing ` id4 `. and protective gaps of this By making use offormula (7), the risk offSlure ofa protected piece of insulationcan be evaluated by means of the lb&wing form& UAS Umu pi(u) [I P.(~~fo(u)~~+~ 0 (v) piO (u)*fo (v) d" . ..( pi(u)* Ri = 0 pP I 15) If the protective gap can be considered`ideal `, the second term in formula (15) is equal to zero, and we obtain the f&ing formuh: GlL8ac Ri = I pi(u)*C1-pP,(u)]f, (W d" . ..(16) 0 The ratio of the 50 percent lightning impulse sparkover voltage to the 50 percent switching impulse sparkovervoltage of a spark gap fan be chosen from a wide range ( 1 + I-5 ) by changing the electrode cxmfiguration. It is, therefore, possible toselect the discharge probability curve of-the gap Pp (U) to switching impulse3 quite irrespective of the discharge probability curve to lightning impulses. Design of a spark gap with a view to switching impulse will make the average expected nu.mber ofa -a. ofxhe gap per year, due to switchi~~~%jG$t% a value .& such as not to make system performme under switching surge signiicantly worse. Therefore, the probability of sparkover of the gap due to switching overvoltages reaching the level evaluated on the assumption that equipment operates as designed ( SW9.3 ) shall be very low. Cousequently, even if the s ark gap behaves like an `ideal' protective no protection against them and iusuldevice, the spark gap will provr& ation shall be designed to withstand this type of switching overvoltages. This is evident from formula ( 14 ) ( cpst of ideal spark gap ). Pi -Pi(U). [I--p(U) J=Ppi(U) . ..(17) As regards switching ovenroltagesuceedin~ the values based on the assumption of correct behaviour of equipmcn~ we can assume that the overvoltage value is such that a gap sparkover will almost certainly be caused and formula ( 11 ) then becomw pi(U) -Pi(U) P*(U) . . . (18) ai 18:3716-En8 and formula (15) u mSx Ri The protective to the relationship those of the gap. It possible to zero over = 0 1 pi (u) pip( u)fo (u)* dU . ..(19) effect is, in this case, due only to Pin ( U) that is, between the time-to-discharge of the insulation and is, therefore, necezsaty to make Pip ( U) as close as the entire range of interest of U. Design of a spark gap with lightning impulses in view will be such as to limit the average expected number of flashovers per year of the gap to lightning surges to an acceptable value Nr. In this connection, it shall bcremembered that in many cases the sparkover of the gap does not lead to any supply interruption. Let us consider the example of a spark gap installed on the line side of the breaker. If the lightning stroke causes the line to flashover, a coincidental sparkover of the gap is of no significance. On the other hand, if the overvoltage amplitude does not reach the sparkover level of the line at the point struck, it is unlikely to cause a sparkover of the gap, even if the gap withstand level islower than that of the line. This is due to the reduction in surge amplitude at the point of installation of the spark gaps, because of attenuation as well as the possible presence of other lines and surge arresters at the station. Thus, in contrast to the case of switching surges, it is possible to accept in certain cases spark gaps with a 50 percent discharge voltage to lightning impulses lower than that obtained, basing the design of the apparatus on switching surg_es and using the combination of the rated impulse withstand voltages glen in Table 3 of IS : 2165-l 977*. Therefore, it may be concluded from formulae 11 and 14 that, spark gaps my offer a limited degree of protection in the case of lightning overvoltages of the order of the rated lightning impulse withstand voltage of the apparatus. Aa regards lightning overvoltages much higher than the rated light&g impulse withstand voltage of the apparatus gap sparkover will almost certainly be caused. In this case formula 18 and the same considerations as regards Pip ( U) previously made in respect of switching surges apply. For air insulation of equipment the condition Pip ( U ) o 0 `may b fulfilled both for lightning and switching impulses by making use of gap l~dation cwxdimtion ( sawd r&rbn ) . h;rvhg a high critical sparkover voltage to switching surges ( kV/cm ), that IS : 3716- 1378 much shorter than air distances of the apparatus, for instance by k gaps making use of spark-gaps having a conductor-rod electrodes. configuration for the For non-self&storing insulation of equipment the check that factor Pip ( U) is reasonably small should be made by means of chopped wave withstand test carried out at a voltage level based on the highest overvoltage which may be expected in the system and with a truncation time to be chosen on the basis of the time-to-sparkover of the gap. These tests are not laid down in IS: 2165-1977* and, if necessary should be agreed upon between the user and the supplier. APPENDIX D ( Clause 9.3 ) EXAMPLES OF APPLICATION D-l. Table 3 illustrates the choice of the insulation level of a piece of equip ment for Case I ( see 9.3 ) protected against both switching and lightning overvoltages by surge arresters mounted at it3 terminals. Example 1 refers to 420 kV transformer and Example 2 to a 765 kV transformer. D-2. The insulation levels of the transformers depend on the protective levels of arresters against both switching and lightning impulses ( Block 7 of Fig. 12 ). The protection level of a particular surge arrester depends, in its turn, on its characteristics and rating. The rating of surge arresters in both Examples 1 and 2 is chosen as the available rating immediately above the temporary overvoltagcs anticipated on the system ( Block 6 ). Temporary overvoltages include voltage rise during faults, overvoltagcs due to inrush currents, sudden load rejection and other causes. The overvoltage to be expected is influenced by the earth fault factor, system configuration, the characteristics of system equipment, and operating practices ' see 3.3 ). D-3. Table 4 shows the choice of the insulation level of a piece of equipment for Case II with no surge arrester protection or with remote surge arrester protection. Example 1 refers to the line-to-earth insulation of a 420 kV disconnecting switch on the line-side of the breaker, no surge arresters being installed at the line entrance. Example 2 refers to a 765 kV disconnecting switch at the same location. D-4. The rated switching iinpulse withstand voltage is selected first from Table 3 of IS: 2165-1977* on the basis of the statistical switching overvoltage level at the equipment location and on the basis of an acceptable risk of failure ( Blocks 4 and 8 of Fig. 12 ). *Insulationco-ordination( scrond WV&I). 83 1) Bo~&Dah Highat voltage for equipment u. (nm) carapoadine line-to-ground voltage rmsvaluc Peak value U.*`T 1 df Detamkdog tempomy overvoltage (derived fram system rtudics ): rmsvaluc Peak value Mium safety fiwtol. Forswitchingovcrvoltagcs. For lightning ovcrvobgcs Rated voltage ( ms) Afar switching impulse sparkover voltage MUXliitnin6 impubc sparkover voltage Mar front-of-wave sparkover vobge MUXresidual voltage at rated di&arge current 3) R&ction Lad: To switching impuka To lightning impulses Ul4lT kV LV kV (p.u.) kV kV ( p.u.) kV kV kV kV kV EXAMPLE I ExAuPU2 420 243 343(1~00) 320 452 ( 1.32 ) 1.15 1.25 324 765 805 1 010 735 765 442 625(1-00) 605 855 ( I.37 ) l-15 1.25 612 1230 1400 1660 1400 1230 (1%) 1440+(2+0) 1415 (2.26) 1425 (2.28) 1.16 kV ( p.u. ) 765 (2.23) kV ( p.u. ) 878*(256) 880 (2%) 950 (2.77) I.24 IntetJur 4) rnndatioa Li?d ( Phase-to-ffirlh ) stiw Min&uin&u~o~eentional switching impulse with- kV @.u. ) Rated switching impulse voltage Ratio of the rated r+c$i~ &pulse withstand ;olge to the swltclumg Impulse pmtectb kV (p.u. ) - Mii conventional lightning impulse withstand v&age Rated light&g impulse withstand voltage Ratio of the rated 1' htuing impulse withstand gge to the ligmiing impulse protective kV (P.u. kV (p.u. - 100 (%20) 1800 (2'76) 1 175 (942) 1.33 1800 (287) 1.25 NOrr - State Rlectricity Boards with 420 kV power system are reqututed to provide the d&aib for Rxample 1 in accordance with the practices they arc follow@ m their *Maximum front-of-Gave sparkover divided by 1.15 ( SW5.2.1). LV Hiiat voltage for qdpm~t Ur ( mu ) 420 765 Comspondiq line-ground voltage: rma value kV 24s 442 Peak vaIue u.~rl~skV ( p.u. ) S4S(l*OO) 625(L.t36j Statistical overvoltage at the equipment location LV ( p.u. ) 910 (2'65) 1'255 (23) ( value exceeded in 2 percent of casa only ) due to rcclaing operationa Maximumacccptcdrisk~6ashoverta~ndfor -10-t 1P recloaingopc!ratioa .:_ statxcal x&ty factor 1.1s 1.24 corrupondingtot8e nurim*-taf I& of flashover ( derived i&n Fig. 7) 1090 Minimum rtatistical awitching imp&c withstand LV 1550 voltage 1050 1550 Ratal switchbg imp&c withstand voltage adectcd kV 4*10-' Risk of Hashoverawnrpanding to tht rated awitch10-b ingimpul~~withstaadvoltageaelectui impulse withstand kV ( p.u. ) 1425 ( 4.15 ) 2 460 (S-84) impulse withstand voita6e --For both ezamples, it is assumed that the only critical switching overvoltages am these due to line re-energization ( that is, it is necessary to check the risk offitihue due to reclosing operations only ). By making use of the simplified statistical approach ( SN 9.2.5 ) and choosing an appropriate standard deviation it is pcssible to determine the statistical safhty factor y corresponding to the mazimum admissible risk of &shover; from factor y it is possible to derive the minimum statistical twitching impulse withstand voltage and then to select the rated switching impulse withstand voltage immediiely above. In the exam* the correlation between risk of ihilure and: ztatistical safktyfdorgiveniaFii.f(~ -8percent)is used. I3-5. Oncethe rated swit&i+mpuJse withstatxd voltage has been obtained a corresponding rated lightning impulse withstand voltage is se&etz~I Eom the same line of Table 8 of IS: 21651977. considering that only the highest value of rated lightning impulse withstand voltage should bc used for apparatus not &ectivdy protected by surge diverters ( SII 6.4 of IS : 216%4977* ). It is then neceszq to verify that this value guarantees a satisfactory under atmospheric overvoltages, that is, a risk of failure not performan%faUua ceudilmtioa (mDndf48iSina). 85 -/ ' U : 3716 - 1978 higher than the permissible one. This can be done in a similar way as previously for switching surges ( Blocks 9 to 12 in Fig. 12 ). In the case under consideration an approximate distribution of the lightning overvoltages, at least ior the case of disconnecting switches in the open position, may be evaluated quite simply on the basis of the line characteristics and of empiricalor semi-empirical laws for wave attenuation. However, lightning overvoltage stresses vary from point to point in a substation. In general, it is therefore extremely difficult and time consuming to achieve the necessary knowledge of stresses by making use of the statistical or simplified statistical approach. Only the highest overvoltage stresses in the most common contingencies ( most common position of breakers and disconnectors ) are therefore evaluated. It is then verified that the rated lightning impulse withstand voltage selected, as said above starting from the rated switching impulse withstand voltage, exceeds the maximum credible atmospheric overvoltages by a suitable margin ( e 10 percent ). If a rated lightning mpulse voltage higher than the one determined on the basis of Table 3 of IS: 2165-1977. is desirable, the value shall be selected from the series in 6.1 ( b) of IS : 2165-1977*. D-6. In Table 5 there is an evaluation of the increase of the risk of failure for the rated switching impulse withstand voltage lower .than the selected value given in Table 3 of IS : 2165-1977*. TABLE S RISK OF FAILURE OF INSULATION AS A FUNCTION OF THE RATBD SWITCHING IMEWLS&~;~m~OLTAGBS OF THE EQUIPUNIT EXAMPLB 1 EXAMPLE 2 1550 (2'48) 1.24 10-r Altamalior (a) -ted switching impulse withstand voltage ( selected value ) Statirtical safety factor Riskof failure corresponding to the above statistical rafety factor Ahmativa (b) Rated switching kV ( p.u. ) 1050 (5'06) 1.16 - 4.10-r 1 950 ( 2.76) 1.05 5 lo-' 1 425 (2.28) 1.14 9lW impulse withstand voltage kV ( p.u. ) Statistical safety factor Risk of failure Af&mativ~ (c) Rated switching kV ( p.u. ) impulse withstand voltage - Statistical safety factor Risk of failure *Insulation co-ordination ( sacond reoisiou ). 1300(2+8) 1.04 &lo-' 86 tS : 3716 - 1978 APPENDIX E ( Clause 9.3 > CLEARANCES IN AIR BETWEEN LIVE CONDUCTIVE PARTS AND EARTHED STRUCTURES TO SECURE A SPECIFIED IMPULSE WITHSTAND VOLTAGE FOR DRY CONDITIONS E-l. In installations which, for various reasons, may not be impulse tested, it is advisable to take steps to avoid flashover occurring below the impulse withstand level which would have been prescribed in the case of a test. E-2. The condition to be fulfilled is that the statistical switching and lightning impulse withstand voltages in air between live parts and earth should be equal to the rated switching and lightning impulse withstand voltage as specified in IS: 2165-1977*. This results in a minimum clearance to be observed which depends on the configuration of the live parts and the nearby structures ( electrode configuration ). E3. No distance is indicated for an equipment which has an impulse included in its specification since compulsory clearances might hamper design of the equipment, increase its cost and impede progress. impulse test, even when only a type test, is sufficient to prove that impulse withstand condition is fulfilled. test the The the E-4. Table 6A and 6B are suitable for general application, providing as a first approximation a clearance to be specified in relation to the insulation 1eveL These tables have been compiled for easy use. E4.1 In Table 6A ( Urn < 245 kV ) reference is made in the first column to the rated lightning impulse withstand voltage and in the second column to the air clearances for unfavourable configurations of live and.earthed parts. E-4.2 In Table 6B ( U, ) 245 kV ) reference is made in the cd (1) and (2) to the values defining the insulation levels and in the co1 (3) and (4) to air-clearances for electrode con8gurations of the ` conductor-structure ' type and ` rod-structure ' type. TM. The c rod-structure ' configuration is the worst electrode configuration normally encountered in practice; the ` conductor-structure ' configuration covers a large range of normally used configurations. In Table 6B reference is made to the electrode configuration because of its notabk intluence for U, > 245 kV. E-6. The values of air clewances given in Table 6 are the minimum values selected by electrical considerations, and do not include any addition tbr construction tolerances, effect of short-circuits, safety of personnel, etc. These values are valid for altitudes not exceeding 1000 m. Quuhtion co-ordination ( saod midon ). 87 (1) LV 45 60 7: 95 125 170 525 380 450 550 650 750 850 950 1050 mm 60 90 120 160 220 520 650 750 900 1100 1500 :5MJ 1660 1900 24aI (1) kV I 750 750 850 850 950 1050 I 175 13lW 1425 1550 (2) kV 850 950 950 1050 1050;1175 1I75;l%KJ;i425 1300;1425;1550 1425 ; 1.590; I &IO 1550;1t?lw;2lfJO 1800;~950;24@0 88 AWUymdmT NO. 2 TO APRIL 1985 IS t 371Q- h7S APPLICATION GUIDE FOR INSULATION CO-ORDI[NATION ( First Revirion ) Altcrodon ( Pw 8, &ace 1.2, fine 1 ) Delete the word ` only `. ezistingandWmmber (P496 8, dam-e 12) - Add the following new clause after the the subsequent clause: 61.3 This guide also covers guidance on phase-to-phase insulation co`nciples and rules for which are enumerated in IS : 2165 ordination the (Part2)-I!383 r (~uAp$endixF).' (PW aistii: 88, A E) - Add the following Appendix aAer the APPENDIX ( Czawc 1.3 ) APPLI%ATION GUIDE F INSULATION FOR PHASETO-PHASE F-X. VotTAoE STRESSES IN SERVICE F-l.8 The dielectric stresseson phase-to-phase insulation may be class&d as follows: a) Power-frequency voltage under normal operating conditions, b) Teqorary overvoltages, c) S~itdaing overvoltages, and d) L&ght&govervoltages. F-1.1 wqw V&age - Under normal operating conditions the power-Gequenq voltage between phases will not be greater than the bighesi v&age for equipment V,. Therefore, LJ, has been taken as a rei&znx value for insulation co-ordination purposes. . pkrx - . : Part 2 Phase-to-phase insulation co-ordiition, princi- F-l.2 Temporary Overvoltages - The clauses of phase-to-phase temporary overvoltage are the same as those of the phase-to-earth temporary overvoltges described in the Standard, except for earth faults which, normally, do not create significant temporary phase-to-phase overvoltages ( less than 1.2 U, ). The insulation performance under temporary overvoltages, which are normally of short duration generally less than one second, is verified by the power-frequency test at voltage levels indicated in Tables 1 and 2 of IS : 2165 ( Part 2 )-1983* for the ranges A and B or specified in certain cases by the relevant Apparatus Committees for range C. F-1.3 Switching Overvoltages - External phase-to-phase insulation is determined by considering the actual stresses due to switching operations, faults and other causes, the behaviour of the insulation when subjected to these stresses and the accepted risk of failure. F-l .3.1 Overvoltages Stresses - The overvoltage between two phases results from two overvoltages to earth, generally having opposite polarity. Experience shows that various combinations of these overvoltages can occur. Nevertheless, the phase-to-phase overvoltage on a three-phase system may be characterized by two situations: Situation 1 : When the peak of the highest phase-to-phase overvoltage is reached on any phase of the three-phase system; Situation 2 : When the peak of the highest overvoltage between phases is attained for any two phases. From these two situations the two phase-to-phase components of the phase-to-phase overvoltage are determined. Whilst, by definition the phase-to=phase earth voltage is always greater in situation 1, it should be noted that the maximum value of the phase-to-phase overvoltage obtained in situation 2 is greater than that obtained in situation 1. The ratio usually observed between the statistical phase-to-phase overvoltage ( situation 2 ) and the statistical phase-to-phase overvoltage ( highest component of situation 1 ) lies in the range of l-5 and 1.8, the values being greater for the higher system voltages. These ratios are the basis for the selection of rated switching-impulse withstand voltage phaseto-phase with respect to the phase-teearth values, F-1.3.2 Insulation Behaviour - The behaviour of any insulation depends on the geometric field distribution and on the type of the dielectric material. In the case of external clearances and for a given total phase-tcr phase test voltage, the division into components using the highest possible l lnaulation co-ordination: and rules. Part 2 Phase-to-phase insulation co-ordinations principles 2 positive component on the one electrode ( and the on the other ) is the most severe condition. negative compotieht However, a different division may be used when testing the insulation provided that the same test severity is achieved by increasing the total phase-to-phase test voltage. The effect of the third phase voltage is generally small and can be neglected in most cases. For the internal insulation, for example, transformers, it seems that the division of the components is of no importance. For other types of insulation, for example, SFs iwulated, three phase enclosed installations the influence of the components has not yet been sufficiently investigated. F-1.3.3 Risk of Failure - Evaluations of the risk of failure between phases have to take into account the division of the phase-to-phase overvoltage into two phase-to-earth components, the statistical distribution of the maximum phaseto-phase overvoltages and the flashover probability of the insulation. Such calculations have been made based on available system data and typical insulation withstand characteristics. The main results show that: a) the two situations mentioned the same risk of failure; and in F.1.3.1 result in approximately b) the risk of failure between phases is smaller than ( or equal to ) that between phase and earth if the phase-to-earth insulation is tested according to IS : 2165 (Part 1)-1977* and if the phase-tophase tested in the same way considering the test procedure given in F-li3.4. F-1.3.4 Test Procedure- To achieve the desired correlation between risks of failure between phases and phase-to-earth the withstand voltage of the insulation has to be equal to or greater than the rated switchingimpulse withstand voltages given in Table 3 of IS : 2165 ( Part 2 ) -1983t. To fulfil this requirement one of the test procedures given in IS : 2165 ( Part 1 )-1977* has to be applied. The permitted number of disruptive discharges includes all discharges between phases and to earth. Among the different division of components which are possible, according to F-1.3.2 the one which is preferred is that giving the same amplitude to both The main reasons are: components ( Uner/UPm - 1 ). a) during the phase-to-phase test the phage-to-phase component shall not be higher than the rated switching-impulse withstand voltage to earth. Taking into account the values given in Table 3 of IS : 2165 ( Part 2 )-1983t this implies a ratio Uneg/UDos already higher than O-73. A further increase of this ratio is necessary to avoid an excessive number of flashovers to earth because of the influence of the negative component applied to the second phase; and rules. *Insulation co-ordination: Part 1 Phase-to-earth iPart 2 Phase-to-phase insulation co-ordination, insulation co-ordination, principles and rules. principles 3 b) for asymmetric test arrangements 4 only two test series, applying the positive component successively to the two phases are required. If the ratio is different from 1, four test series are necessary to cover all cases by permuting the polarities and the values; and the influence of surrounding objects during the test is kept to a minimum. The selection of the ratio U&U, = 1 results in a maximum acceptable risk of faihue between phases which is lower than the maximum acceptable risk of failure to earth. F-l.4 Lightning Overvoltages - When a direct lightning stroke occurs to a phase conductor or a backflashover takes place the lightning stress between phases does not normally exceed the lightning stress to earth. F-2. CLEARANCE IN AIR BETWEEN PHASES TO ASSURE A `SPECIFIED IMPULSE WITHSTAND VOLTAGE IN INSTALLATIONS F-2.1 In complete installations ( for example substation ) which cannot be' impulse-tested as a whole, it is necessary to ensure that the impulse strength is adequate. The statistical switching- and lightning-impulse withstand voltages for phase-to-phase insulation in air should be equal to, or greater than, the rated switching- and lightning-impulse withstand voltages as specified in this standard: Following this principle minimum clearances have been determined for different electrode configurations. Tables 7 and 8 are suitable for general application, as they provide a specified minimum clearance in relation to the insulation level. These clearances may be lower if it has been proved by tests on actual or similar configurations that the required rated impulse withstand voltages are f%lfilled, taking into account all relevant environmental conditions which can create irregularities on the surface of the phase electrode, for example, rain, pollution. As regards range C, lower clearances may also be used if it has been confirmed by operating experience that the switching overvoltages are lower than the rated values indicated in Table 3 of IS : 2165 ( Part 2 )1983* for a given voltage U,. No distance is indicated for an equipment which has a phase-tophase impulse test included in tke specificaation, since mandatory dearanca might hamper the design of the equipment, increase its cost and impede progress. In Table 7 ( U, < 309kV ), first column, the rated lightningimpulse withstand voltages are listed. The second column lists air clearutd ruler. *Insulation co-ordinntion:Part 2 Phase-to-phore insulation co-ordination, principles antes for u&voumbIe con6gurations of energized parts with a rehtively smah curvature radius. These clearances have been derived by the testing procedure desc&ed in 8 of IS : 2165 ( Part 2 )-1983*. In the first cohmm of Table 8 ( U,,, > 300 kV ) rated switchingimpulse withstand voltages phase-*phase are listed. The second column lists clearances for a conductor-conductor ( parallel ) configuration having a symmetrical gap geometry. The same clearances may also be used for other configurations with symmetrical gap geometry such as crossed conductors ofrod-rod gape. The third cohrmn refers to a configuration such as rod-conductor having an asymmetrical gap geometry. For ring-ring gaps or configurations with large smooth electrodes having a higher degree of f&d homogeneity, lower clearances than those given in the second cohunn may be used, provided that the influence of the environmental conditions ( AWabove ) is taken into account. The cIearancesfor the phase-to-phase insulation ( SMTables 7 and 8 ) can be applied together with the clearances for the phase-to-earth insulation. `i%BI.R7 CO-%7ONSBR'lWEEN~~NYAIUD IarNamuPHAsR-To-PRASRAIR -PORUm<=kV (CkllfCCIW'I) RATBDLIQETHIX~IMPULS~ WITEBTARDVOLTAOL, PHASE-T~-PEAEQS (P-X) (1) LV $ 6750 95 125 145 170 z 450 so 650 750 850 950 lO!X-l &mUUX hA86+tbP~ASB hIpcmABuocIIB (21 mm 2 1% 160 5 z 900 1 100 :z 1700 .:z lInn1l8tioncowrdiiation: Put 2 Phase-to-phaskinsulation co-ordiiation principla mdNle& 5 The values of air .clearances given in Table 7 and 8 are dictated by dielectric considerations. Other factors such as construction tolerances, the effect of short circuits, wind, safety of personnel, maintenance, corona effects etc. are not included. The indicated values are valid for altitudes not exceeding 1 000 m. The effects of higher altitudes are under consideration. TABLE 8 CORRELATION BETWEEN INEULATION LBVBLS AND MINIMUM PHASE-TO-PHASE AIR CLEABANCES FOR UEJ > 300 kV (Chn.r~ F-2.1 ) RATED .?iWXTO~X?@IY~ULSBWITHSTA~ V~LTAGUCPHA(IE-TO-PHAEJB (fi=) (1) iV 1 175 1300 1 425 1550 1 675 :k% 2 loo 2250 2400 2550 M~~~J~PEA~~~o-PEAsEAIB hlUE~CZlrOB~NIIOUBATIOI?S -1 c Conduc;to$ductor Rod-Conductor 12) m 2'4 ;:; x: 4'3 f; (3) m , (ETDC19)